United States v. Slough

MEMORANDUM OPINION

Granting the Defendants’ Motion to Dismiss the Indictment; Denying as Moot the Government’s Motion to Dismiss the Indictment Against Defendant Slatten Without Prejudice

[T]he basic purposes that lie behind the privilege against self-incrimination do not relate to protecting the innocent from conviction, but rather to preserving the integrity of a judicial system in which even the guilty are not to be convicted unless the prosecution shoulder the entire load. 1

*115 I. INTRODUCTION

The defendants have been charged with voluntary manslaughter and firearms violations arising out of a shooting that occurred in Baghdad, Iraq on September 16, 2007. They contend that in the course of this prosecution, the government violated their constitutional rights by utilizing statements they made to Department of State investigators, which were compelled under a threat of job loss. The government has acknowledged that many of these statements qualify as compelled statements under Garrity v. New Jersey, 385 U.S. 493, 87 S.Ct. 616, 17 L.Ed.2d 562 (1967), which held that the Fifth Amendment privilege against self-incrimination bars the government from using statements compelled under a threat of job loss in a subsequent criminal prosecution. The Fifth Amendment automatically confers use and derivative use immunity on statements compelled under Garrity; this means that in seeking an indictment from a grand jury or a conviction at trial, the government is prohibited from using such compelled statements or any evidence obtained as a result of those statements.

The government has also acknowledged that its investigators, prosecutors and key witnesses were exposed to (and, indeed, aggressively sought out) many of the statements given by the defendants to State Department investigators. Under the binding precedent of the Supreme Court in Kastigar v. United States, 406 U.S. 441, 92 S.Ct. 1653, 32 L.Ed.2d 212 (1972) and this Circuit in United States v. North, 910 F.2d 843 (D.C.Cir.1990), the burden fell to the government to prove that it made no use whatsoever of these immunized statements or that any such use was harmless beyond any reasonable doubt.

Beginning on October 14, 2009, this court convened a Kastigar hearing to explore whether the government had made any use of compelled statements during its prosecution of the defendants. During this hearing, which spanned three weeks, the parties presented testimony from twenty-five witnesses, including the government’s entire prosecution team, the lead FBI agents in charge of the investigation and all five defendants. The parties offered hundreds of exhibits into evidence and submitted voluminous pre- and post-hearing memoranda.

From this extensive presentation of evidence and argument, the following conclusions ineluctably emerge. In their zeal to bring charges against the defendants in this case, the prosecutors and investigators aggressively sought out statements the defendants had been compelled to make to government investigators in the immediate aftermath of the shooting and in the subsequent investigation. In so doing, the government’s trial team repeatedly disregarded the warnings of experienced, senior prosecutors, assigned to the case specifically to advise the trial team on Garrity and Kastigar issues, that this course of action threatened the viability of the prosecution. The government used the defendants’ compelled statements to guide its charging decisions, to formulate its theory of the case, to develop investigatory leads and, ultimately, to obtain the indictment in this case. The government’s key witnesses immersed themselves in the defendants’ compelled statements, and the evidence adduced at the Kastigar hearing plainly demonstrated that these compelled statements shaped portions of the witnesses’ testimony to the indicting grand jury. 2 The explanations offered by the *116 prosecutors and investigators in an attempt to justify their actions and persuade the court that they did not use the defendants’ compelled testimony were all too often contradictory, unbelievable and lacking in credibility.

In short, the government has utterly failed to prove that it made no impermissible use of the defendants’ statements or that such use was harmless beyond a reasonable doubt. Accordingly, the court must dismiss the indictment against all of the defendants.

II. BACKGROUND

A. Factual Background

1. The Nisur Square Incident

The defendants — Paul Slough, Evan Liberty, Dustin Heard, Donald Ball and Nicholas Slatten — were security guards employed by Blackwater Worldwide (“Blackwater”), a private company that provided security services to U.S. government employees operating in Iraq. Govt’s Pre-Hr’g Mem. at 2. On September 16, 2007, the defendants were part of a Black-water Tactical Support Team answering to the call sign “Raven 23,” whose function was to provide back-up fire support for other Blackwater personal security details operating in Baghdad. Id.

The Raven 23 convoy consisted of four vehicles. Govt’s Ex. 2. 3 Defendant Ball functioned as the rear turret gunner on the lead vehicle. Id. Defendants Liberty, Slough and Slatten were positioned in the third vehicle as the driver, turret gunner and designated defensive marksman (or sniper) respectively. Id. Defendant Heard was the rear turret gunner in the fourth vehicle. Id. Jeremy Ridgeway, who pleaded guilty to charges stemming from the incident and has been cooperating with the government, functioned as the lead turret gunner in the fourth vehicle. Id. The defendants were armed with machine guns, grenade launchers, rifles and pistols. Govt’s Post-Hr’g Mem. (“Govt’s Mem.”) at 2.

Shortly before noon on September 16, 2007, Raven 23 received a message that a vehicle-borne improvised explosive device (“VBIED”) had detonated in the vicinity of a compound in which U.S. officials were meeting with Iraqi officials. Defs.’ Mot. for an Evidentiary Hr’g Under Garrity and Kastigar at 4. The Raven 23 convoy subsequently took up positions in Nisur Square, a traffic circle located just outside the International Zone in downtown Baghdad, to secure an evacuation route for the American officials and the Blackwater team providing their security. Id. Soon after the Raven 23 vehicles entered the traffic circle, a shooting incident erupted, during which the defendants allegedly shot and killed fourteen persons and wounded twenty others. Govt’s Mem. at 2. The government contends that the dead and wounded were unarmed civilians who were the victims of unprovoked violence by the defendants. The defendants maintain that they came under attack by insurgents and that their actions constituted a legitimate response to a mortal threat.

The paroxysm of violence that occurred on September 16, 2007 in Nisur Square triggered immediate responses that would have far-reaching consequences. As discussed below, the State Department, the U.S. military and Iraqi forces commenced immediate inquiries into the shooting. *117 Media began investigating the incident, interviewing eyewitnesses and probing sources in the U.S. government. And ultimately, the decision was made to commence a criminal prosecution against the defendants in this case.

2. The Defendants’ September 16, 2007 Statements to State Department Investigators

Hours after the shooting, the Department of State’s Diplomatic Security Service (“DSS”) directed all the members of the Raven 23 convoy to submit to interviews at the State Department offices in Baghdad, which are referred to by U.S. personnel as “the Palace.” Hr’g Tr., Oct. 14, 2009 p.m. at 20; 4 Hr’g Tr., Oct. 19, 2009 a.m. at 99-100; Defs.’ PosNHr’g Mem. (“Defs.’ Mem.”) at 5; Govt’s Mem. at 3. DSS Special Agent Theodore Carpenter oversaw these interviews, see Govt’s Mem. at 5, which were conducted by DSS Special Agents Michael Scollan, Lisa Lopez, Mario Reta and David Motley, see Defs.’ Mem. at 6-7.

The September 16, 2007 interviews delved into the particulars of the Nisur Square shooting. The defendants and the other members of the Raven 23 convoy provided the interviewing DSS agents with detailed accounts of the actions they purportedly took at Nisur Square. From these accounts, the DSS agents subsequently prepared two “Memorandum Reports of Interviews” that memorialized these oral statements. See generally Defs.’ Reta Ex. 3 5 (Mem. Report of Interview, Sept. 16, 2007); Defs.’ Lopez Ex. 3 (Mem. Report of Interview, Sept. 16, 2007).

During the September 16, 2007 interviews, four of the five defendants acknowledged that they had fired their weapons at Nisur Square. Defendant Slough, the turret gunner in the third vehicle in the convoy, reportedly stated to DSS agents that

[a] white vehicle approached the team at a high rate of speed and would not stop despite his hand signals and throwing a water bottle. Other civilians tried to wave the vehicle down, but it still would not stop. [Slough] engaged and hit the driver. An Iraqi Policeman, wearing [a] blue button down shirt and black pants, began to push the vehicle towards [the] team. [Slough] engaged [the] vehicle a second time and [the] Iraqi Policeman ran away. [Slough] then witnessed muzzle flashes from a shack, returned fire, and hit the individual.

Defs.’ Lopez Ex. 3 at 1.

Defendant Slatten, also stationed in the third vehicle, reportedly told the DSS agents that the Raven 23 convoy began “receiving small arms fire from the left side of their vehicles by two individuals in the tree line. [Slatten] returned fire with two rounds hitting one of the individuals.” Id. Defendant Heard, positioned in the rear vehicle, stated that he “fired on the white vehicle that refused to stop with approximately one full magazine from [his] M-4. He then saw muzzle flashes from an individual several meters behind [the] white vehicle and engaged [that individual] with five to ten rounds from [his] M-4. He continued to receive fire from the individual and responded with one round from his *118 203.” 6 Id. Defendant Ball acknowledged to DSS Agents Scollan and Reta that he fired two rounds into the driver’s door of the white vehicle engaged by Slough and Heard. Defs.’ Reta Ex. 3 at 2.

Defendant Liberty did not acknowledge firing his weapon during his September 16, 2007 interview. See id. at 3-4. He did, however, disclose that the convoy “started receiving small arms fire from the six o’clock position,” that he “saw two to three Iraqi Police officers with AK-47s shooting at Raven 23 members,” and that he “saw one Iraqi Police combatant neutralized at the guard shack, approximately 25 meters away, and a second neutralized in the south-east corner of the traffic circle.” Id. He specified that “most of the hostile activity toward Raven-23 ... came from his 7 to 9 o’clock position.” Id.

3. The Hunter Memorandum

The September 16, 2007 interviews appear to have been conducted in connection with the reporting procedures set forth in a State Department memorandum authored by Regional Security Officer Mark Hunter and titled “WPPS [7] On-Duty Discharge of Firearm Reporting Procedures” (“the Hunter Memorandum”). See Govt’s Ex. 32 at 1-2; Govt’s Mem. at 4. The Hunter Memorandum required all Black-water personnel involved in a shooting incident to report immediately to the Regional Security Office Tactical Operations Center for a debriefing by State Department officials. Govt’s Ex. 32 at 1. After the debriefing, any employee who discharged his weapon was to be given a form, a template for which was attached to the Hunter Memorandum, on which to prepare a written statement. Id. The form, which bore the heading “Sworn Statement,” provided as follows:

I,-, hereby make the following statement at the request of _, who has been identified to me as a Special Agent of the U.S. Department of State, Diplomatic Security Service. I understand that this statement is made in furtherance of an official administrative inquiry regarding potential misconduct or improper performance of official duties and that disciplinary action, including dismissal from the Department’s Worldwide Personnel Protective Services contract, may be undertaken if I refuse to provide this statement or fail to do so fully and truthfully. I further understand that neither my statements nor any information or evidence gained by reason of my statements can be used against me in a criminal proceeding, except that if I knowingly and willfully provide false statements or information, I may be criminally prosecuted for that action under 18 United States Code, Section 1001.

Id. at 3 (emphasis added). This admonishment — that an employee must make a statement or face termination but that any statement so made cannot be used in a subsequent criminal prosecution — is commonly referred to as a “Garrity warning” or “Kalkines warning.” See Garrity, 385 U.S. at 500, 87 S.Ct. 616; Kalkines v. United States, 200 Ct.Cl. 570, 473 F.2d 1391, 1393 (1973) (holding that prosecutors may not use statements compelled under a threat of removal from office or use any information derived from those statements). Prior to the Nisur Square shoot *119 ing, all of the defendants had been involved in previous shooting incidents, after which they had been required to provide a sworn written statement on the form described above. See Hr’g Tr., Oct. 27, 2009 a.m. at 10-11; Hr’g Tr., Nov. 3, 2009 p.m. at 49-50, 82-83,117,136-37.

4. The Defendants’ September 18, 2007 Written Statements and the State Department’s Continuing Investigation

On September 18, 2007, the defendants and their fellow Raven 23 team members submitted written statements regarding their actions at Nisur Square to the State Department on the Hunter Memorandum form. 8 Defs.’ Hulser Ex. 17 at 3 (Letter from Prosecutor Joseph Raster to Defense Counsel, June 30, 2009) (“Govt’s June 30, 2009 Letter”). The defendants’ September 18 written statements expanded on the oral statements they gave to DSS investigators on September 16. See generally Slough Stmt. (Sept. 18, 2007); Slatten Stmt. (Sept. 18, 2007); Ball Stmt. (Sept. 18, 2007); Heard Stmt. (Sept. 18, 2007); Liberty Stmt. (Sept. 18, 2007); see also Defs.’ Farrington Ex. 4.

On September 18, 2007, at the Iraqi National Police (“INP”) Headquarters, DSS investigators also interviewed fourteen Iraqi nationals who claimed to have witnessed the Nisur Square incident. Defs.’ Carpenter Ex. 8. The interviewees were identified by the INP and many were INP officers or traffic policemen posted at Nisur Square at the time of the incident. Id. In a memorandum summarizing these interviews, Agent Carpenter stated that “[s]ome of the questions asked of these witnesses may have been formed as a result of information provided by [Raven] 23, but was limited due to not wanting to divulge what we had been told by [Raven] 23 and to just get the witnesses’ observations of the incident.” Id.

Over the following days, defendants Slough, Slatten, Ball and Heard submitted to additional questioning by DSS agents regarding their actions at Nisur Square. See Defs.’ Carpenter Ex. 9; Defs.’ Lopez Ex. 5; Govt’s Mem. at 6. On September 20, 2007, several DSS investigators, including Agent Carpenter, interviewed these four defendants in an effort to focus on “the specific details” of the Nisur Square incident in light of the sworn statements submitted two days earlier. See Defs.’ Carpenter Ex. 9. Later that day, DSS investigators conducted an incident scene investigation at the Nisur Square traffic circle. Defs.’ Carpenter Ex. 8. DSS investigators took photographs, interviewed additional witness and collected physical evidence for analysis. Id.

On September 23, 2007, Agents Carpenter and Scollan re-interviewed Slough “to clarify details relating to the threat types and locations engaged by Slough, weapons used to engage the threats, and the number of rounds fired at each threat.” See Defs.’ Scollan Ex. 4 (Mem. Report of Interview, Sept. 23, 2007) at 1. The interview report states that at the outset of the interview, “Carpenter admonished Slough of the Ealkine warning concerning the administrative nature of the investigation, and reminded him, that should he provide a false statement, he could be criminally prosecuted for giving the false statement.” Id. During this interview, Slough acknowledged for the first time that he had fired an M203 grenade at the white sedan. Id. at 2.

*120 On September 24, 2007, DSS agents conducted additional re-interviews of Slough, Heard and Ball to specifically address the number of grenades fired during the Nisur Square incident. See Defs.’ Scollan Ex. 5. During these interviews, the defendants were shown photographs of the scene as well as aerial photographs of the traffic circle and were asked to specify the directions in which they had fired their weapons and the locations of the threats they had allegedly seen. Id.

5. Dissemination of the Defendants’ Statements Made to DSS Investigators

Because of the nature of use immunity, Garrity statements may present thorny legal issues, even when the statement, itself, has not been introduced at trial or even seen by the prosecution team. This is true because an internal investigator, or other person having access to the Garrity statement, may reveal its contents to a federal investigator, prosecution witness, or the media — some times without revealing its source. When this happens, it is possible that the testimony of other witnesses may become contaminated. 9

Media reports regarding the Nisur Square shooting began to appear almost immediately after the incident. Several of these early reports included statements from State Department officials asserting that Raven 23 had been engaged in a firefight at Nisur Square and that its actions were taken to neutralize various insurgent threats. See, e.g., Defs.’ Media Exs. 22, 23, 24, 25. 10 Some of these articles made specific reference to the fact that some Raven 23 team members claimed that they had received incoming fire. See Defs.’ Media Ex. 28.

In the following days, the defendants’ September 18 written statements, which the government has conceded were compelled under Garrity, were leaked to the media and disseminated globally in news reports. See, e.g., Defs.’ Media Exs. 2, 6, 7. The earliest of these reports was a September 28, 2007 article in which ABC News reported that it had obtained “sworn statements given to State Department investigators” in which Blackwater personnel described the events that had occurred at Nisur Square. Defs.’ Media Ex. 2. The report quoted from the defendants’ September 18 written statements. Id. Later, the Washington Post published an article contrasting the accounts of the four Raven 23 team members with those of Iraqi eyewitnesses. Defs.’ Media Ex. 13. On November 14, 2007, ABC News posted Slough’s September 18 written statement, in its entirety, on its website. Defs.’ Media Ex. 7. Ultimately, portions of the defendants’ September 18 statements appeared in reports by ABC News (including in a nationally broadcast television segment), MSNBC/Newsweek, the Washington Post, the New York Times and Time/ CNN, among other sources. See Defs.’ Media Exs. 2, 6, 7,13,14, 23.

Press accounts regarding the Nisur Square incident pervaded the national news in both Iraq and the United States for two weeks. Through these reports, Adam Frost, Matthew Murphy and Mark Mealy — members of the Raven 23 convoy and key witnesses for the prosecution— were repeatedly exposed to the defendants’ September 18 written statements. *121 Frost and Murphy both acknowledged reading the September 18 statements of Slatten and Slough, Hr’g Tr., Oct. 21, 2009 p.m. at 25-26; Hr’g Tr., Oct. 14, 2009 p.m. at 6, and seeing countless news articles regarding the incident through their “Google News” alerts, many of which referenced and quoted the defendants’ sworn statements, Hr’g Tr., Oct. 21, 2009 p.m. at 26-27; Hr’g Tr., Oct. 14, 2009 p.m. at 27. Likewise, Mealy acknowledged that he may have read the sworn written statements of other Raven 23 team members. Hr’g Tr., Oct. 19, 2009 a.m. at 54.

6. The Investigations by Iraqi Forces, the U.S. Military and the Prosecution

Immediately after the Raven 23 convoy left Nisur Square on September 16, 2007, INP officers entered the traffic circle and commenced an investigation. Govt’s Mem. at 7. These officers began collecting physical evidence, including shell casings and spent ammunition magazines, one of which had the name “Liberty” written on it. Id. In addition, Iraqi police interviewed victims and eyewitnesses, including fellow INP officers. Id.

Within thirty minutes of the shootings, U.S. Army Colonel David Boslego, who had been responsible for training the INP and was present at INP Headquarters, also arrived at the Nisur Square traffic circle. Id. Colonel Boslego examined the scene for evidence of an insurgent attack. Id. Although he found no evidence of insurgent activity, he discovered numerous American shell casings, a spent M203 grenade cartridge and two spent magazines bearing the name “Liberty.” Id. at 7-8.

Another high-ranking American officer, Colonel Michael Tarsa, arrived to inspect the scene within an hour after the incident. Id. at 8. Colonel Tarsa was the battalion commander responsible for operations in the area of Baghdad that included Nisur Square. Hr’g Tr., Oct. 21, 2009 a.m. at 13-14. Like Colonel Boslego, Colonel Tarsa observed numerous shell casings but found no evidence of an insurgent attack. Id. at 25-26.

In early October 2007, FBI investigators arrived in Baghdad to investigate the Nisur Square shooting. Govt’s Ex. 207. Over the next several weeks, FBI investigators interviewed over seventy-five Iraqi witnesses and victims identified by the INP. Govt’s Mem. at 7. The FBI also conducted dozens of interviews of Black-water personnel, U.S. military police and INP officers. Hr’g Tr., Oct. 21, 2009 p.m. 102-03. FBI investigators collected physical evidence from the scene, such as shell casings, bullet fragments and weapons, and examined the vehicles allegedly fired on by the defendants, the Raven 23 vehicles and the defendants’ weapons. Govt’s Mem. at 9.

Among the Blackwater personnel interviewed were members of the Raven 23 convoy still present in Iraq. Hr’g Tr., Oct. 21, 2009 p.m. at 103-04. Three of the individuals interviewed — Adam Frost, Matt Murphy and Mark Mealy- — had raised concerns about the events at Nisur Square with their Blackwater supervisors in the immediate aftermath of the shooting. Hr’g Tr., Oct. 14, 2009 a.m. at 34-35. All three expressed a willingness to cooperate with the government’s investigation. Hr’g Tr., Oct. 21, 2009 p.m. at 114; Hr’g Tr., Oct. 19, 2009 a.m. at 26; Govt’s Ex. 207 at 11.

7. Hulser’s Cautionary Intervention

A defendant’s rights under Garrity are protected, in part, by self-policing by government prosecutors. This happens, in large measure, by prosecutors having potential Garrity materials reviewed *122 and redacted by persons who are not part of the prosecution team. 11

On September 26, 2007, representatives of the FBI and the Department of Justice (“DOJ”) Criminal Division met with State Department representatives to discuss the preliminary findings of the DSS investigation into the Nisur Square shooting. Defs.’ Mot. for Evidentiary Hr’g Under Kastigar and Garrity, Ex. 23 (Letter from Prosecutor Joseph Raster to Defense Counsel, May 29, 2009) (“Govt’s May 29, 2009 Letter”) at 2. At that meeting, State Department representatives distributed copies of a DSS report that summarized the State Department’s initial findings, including information gained from the oral and written statements provided by the defendants to DSS investigators. Id.

On September 28, 2007, the State Department’s Office of Legal Counsel contacted the DOJ’s Criminal Division and expressed concern that some of the information used to prepare the DSS report may have been based on compelled statements made by Blackwater personnel. Id. In light of these concerns, the government assigned to the case a team of prosecutors and investigators who had not participated in the September 26, 2007 meeting and had not been exposed to the information discussed at that meeting. Id. The government barred the DOJ Criminal Division from further involvement in the case and assigned prosecutors in the Counterterrorism Section of the National Security Division (and later the United States Attorney’s Office for the District of Columbia) to investigate the Nisur Square shooting. Id.; Hr’g Tr., Oct. 22, 2009 p.m. at 78. Assistant United States Attorney Kenneth Kohl was assigned to lead the new trial team, to be assisted by Department of Justice trial attorney Stephen Ponticello. 12 Defs.’ Mem. at 26. FBI Special Agent John Patarini was assigned as the team leader and case agent for the investigative team. Hr’g Tr., Oct. 21, 2009 p.m. at 63-64.

In early October 2007, Raymond Hulser, a Deputy Chief in the Public Integrity Section of the Criminal Division, was assigned as the “taint attorney” for the Nisur Square investigation. Govt’s May 29, 2009 Letter at 3. He was selected for that role because of his experience with Garrity and Kastigar issues. Hr’g Tr., Oct. 23, 2009 a.m. at 17-18. As head of the “taint team,” Hulser was charged with “prevent[ing] the prosecutors and investigators who were to handle the investigation from being exposed to potentially compelled statements or information derived from such statements” by pre-screening information obtained from sources including the State Department, Blackwater and the media, prior to the disclosure of such information to the trial team. Govt’s May 29, 2009 Letter at 3. In addition, in October 2007, Hulser and Jesse Tampio, a State Department attorney, drafted a written protocol to govern the use of information derived from DSS personnel in the course of the investigation. Id.; see Defs.’Hulser Ex. 2. The protocol was designed to prevent the FBI investigative team operating in Baghdad from exposure to information derived from any potentially compelled statements obtained by DSS investigators. Govt’s Mem. at 6; Defs.’ Hulser Ex. 2.

8. Failure of the Government’s Taint Procedures

Unless the government relies solely upon evidence obtained prior to the im *123 munized testimony, the principles of Kastigar generally require (as a practical matter) a showing that prosecuting officials and their agents were aware of the immunity problem and followed reliable procedures for segregating the immunized testimony and its fruits from officials pursuing any subsequent investigations. 13

By all accounts, these prophylactic measures fell well short of expectations, particularly with respect to the defendants’ September 16, 2007 oral statements to DSS investigators. Throughout the investigation, Hulser repeatedly advised the trial team not to obtain or rely on any information derived from the defendants’ September 16 oral interview statements. For instance, on November 29, 2007, Hulser wrote an e-mail to Michael Mullaney, a Section Chief within the DOJ’s National Security Division, 14 in which he stated that “[w]hile no Kalkines warning was given before [the September 16, 2007] interviews, I believe that we should treat all of their statements to the [DSS agents] as compelled given the practice of routinely giving the Kalkines warning to the participants in a shooting.” Defs.’ Kohl Ex. 10. Mullaney forwarded Hulser’s e-mail to Kohl, who responded, “Got it. Thanks Mike.” 15 Id.

On December 10, 2007, Hulser wrote to Mullaney expressing concern about the trial team’s intention to interview Agent Carpenter and other DSS agents who had interviewed the defendants and participated in the Nisur Square investigation. See Govt’s Ex. 57 at 1. As Hulser explained, “[g]iven that they took, reviewed and/or wrote a report based on compelled statements, they are thoroughly tainted, and even the focus of the investigative steps that they took would have been affected by the compelled statements.” Id. (emphasis added).

On January 15, 2008, Hulser again wrote to Mullaney, advising him that he viewed “any statements taken after the 9/16 incident whether they occurred before or after the sworn, written statements” as potentially compelled. Defs.’ Hulser Ex. 10. Hulser reasoned that “[g]iven the routine practice of getting compelled written statements following shooting incidents, and given that the Team 23 members had themselves provided compelled statements following prior incidents, the Team 23 members may claim that they thought all of their post-shooting statements and interviews were compelled, regardless of the form in which they were taken or the order in which they provided them.” Id. Hulser then reiterated that

the FBI needs to understand that in interviewing Ted Carpenter, who was involved in the initial investigation and who is thoroughly tainted, there is a substantial risk that the FBI will be tainted. It is not only Ted’s written interview reports and notes that contain tainted materials; rather, his thought processes, the focus of his investigation, what he looked at and when — these may *124 all have been influenced by compelled statements. We’ve got an uphill battle on this Garrity issue, and the burden of proof is ours, so we need to be particularly cautious.

Id. (emphasis added).

Likewise, on February 4, 2008, Hulser wrote an e-mail to Mullaney stating that he believed “any statement given by the 9/16 participants following the incident, whether in the form of interviews, after-action-reports, or written statements with Kalkines warnings should be treated as compelled because of the pervasive practice of compelling them to give statements after every incident.” Defs.’ Hulser Ex. 6. In a separate e-mail also sent on February 4, 2008, Hulser advised Mullaney that he did not plan to provide the trial team with any reports of Agent Lopez’s interviews of the defendants “or any summary or spot reports, as those materials may contain information from compelled (or arguably compelled) statements of the individuals who remain subjects” of the investigation. 16 Defs.’ Kohl Ex. 16.

In direct contravention of Hulser’s unequivocal warnings, in January and February 2008, the government’s trial team interviewed all of the DSS agents who had conducted the September 16, 2007 interviews and specifically inquired about the details of the defendants’ statements during those interviews. Defs.’ Mem. at 28; Govt’s Mem. at 13-14. For instance, on January 11, 2008, the trial team interviewed Agent Carpenter, whom Hulser had specifically described as “thoroughly tainted,” Defs.’ Hulser Ex. 10, and obtained information regarding the defendants’ statements to DSS investigators, Hr’g Tr., Oct. 19, 2009 a.m. at 30. Two weeks later, the trial team interviewed Agent Lopez and obtained her interview notes from the interviews conducted on September 16, 2007. Hr’g Tr., Oct. 15, 2009 p.m. at 60-61. Indeed, Kohl testified that both he and his investigators were exposed to a tremendous amount of information from the defendants’ September 16, 2007 interview statements. 17 Hr’g Tr., Oct. 29, 2009 a.m. at 115.

Furthermore, in February 2008, prosecutors used the defendants’ September 16, 2007 statements to obtain a search warrant authorizing the government to search the e-mail accounts of the Raven 23 team members to obtain unsigned drafts of their September 18, 2007 written statements. 18 See Defs.’ Mem. at 31-32; see generally Govt’s Ex. 288. To demonstrate probable cause justifying the search, Agent Patarini noted in his affidavit that “within hours after the incident, all 19 members of [Raven 23] were interviewed by agents of the [DSS]” and that “[t]hese interviews confirmed that only a few of the guards actually fired them weapons.” Govt’s Ex. 288 at 8. Hulser was never informed about the search warrant and never approved the use of the September 16 interview state *125 ments to obtain the search warrant for the September 18 written statements. 19 Hr’g Tr., Oct. 23, 2009 a.m. at 28-29. Hulser testified that had he been informed, he would not have permitted the trial team to view any records obtained from the search prior to screening by a taint attorney. Id.

Agent Patarini executed the search warrant on February 15, 2008 and obtained drafts of the September 18 written statements of defendants Ball and Slatten. Hr’g Tr., Oct. 22, 2009 p.m. at 14-17. Agent Patarini reviewed the draft statements in the spring or summer of 2008 and included those statements in a binder of search warrant materials he provided to the prosecution team. Id. at 17-22; Hr’g Tr., Oct. 28, 2009 p.m. at 46. Kohl testified that he admonished Agent Patarini for reviewing the draft statements and directed him not to disclose the contents of those statements. Hr’g Tr., Oct. 28, 2009 p.m. at 46. However, Agent Patarini testified on this point and before the court that the search warrant was “a project driven by [Kohl].” Hr’g Tr., Oct. 22, 2009 p.m. at 23.

The divergent views of Hulser and Kohl regarding the status and viability of the defendants’ September 16 interview statements came to the fore in April 2008. On April 11, 2008, Kohl wrote to Hulser expressing frustration at Hulser’s reluctance to approve the trial team’s use of “incident reports,” given his understanding that Hulser had “previously approved and cleared other materials that contain statements made by the targets of our investigation, on the theory that the reports and notes were generated prior to the existence of the sworn Kalkines statements made by these individuals.” Defs.’ Hulser Ex. 8. Hulser responded that he had never approved the use of such statements by the defendants. Id. 20 Three days later, Hulser wrote to Kohl reiterating that his “concern about the immediate debriefs and reports [was] that they were just one part of a mandatory and compelled process to get information after the shooting incidents.” Defs.’ Hulser Ex. 9.

On April 18, 2008, Kohl and Hulser met to discuss their disagreement over the taint issue. Hr’g Tr., Oct. 29, 2009 a.m. at 78; see Defs.’ Kohl Ex. 21. At the conclusion of this meeting, Kohl and Hulser agreed that for all future requests for information from the prosecution team, Kohl alone would determine what investigative materials should be provided directly to the trial team and what materials should be provided to Hulser for screening. See Defs.’ Kohl Ex. 21.

Despite these efforts, “miscommunieations” — as the government charitably deems them — persisted even after the April 18, 2008 meeting. For instance, Kohl testified that during the meeting, Hulser “said he wished [the trial team] could have delayed as long as possible before [it was] exposed to” the defendants’ September 16 interview statements, but that he did not expressly prohibit the trial team from making use of those statements. 21 Hr’ Tr., Oct. 29, 2009 a.m. at 79. Hulser, on the other hand, testified that during the April 18, 2008 meeting, he ad *126 vised Kohl to make no use whatsoever of the defendants’ September 16 interview statements already in his possession. Hr’g Tr., Oct. 23, 2009 a.m. at 42.

In July 2008, Kohl approached counsel for defendant Heard to discuss whether Heard would be willing to cooperate with the government’s investigation. Hr’g Tr., Oct. 29, 2009 a.m. at 161-62. During that meeting, Kohl referenced statements made by Heard during his September 16 interview — specifically, his failure to disclose firing an M203 grenade during his initial interview followed by his admission to DSS investigators later that evening that he had, in fact, fired a grenade. Id. at 162-63. Kohl testified that he raised this fact during the plea discussions because he viewed it as evidence reflecting Heard’s consciousness of guilt. Id. at 163.

Furthermore, in August 2008, Kohl requested and obtained information regarding the defendants’ September 16 interview statements to DSS investigators— namely, the two “Memorandum Reports of Interviews” prepared by the interviewing DSS agents. Hr’g Tr., Oct. 29, 2009 a.m. at 165-66; Govt’s June 30, 2009 Letter at 4. Kohl read these documents in the course of considering whether a false statements charge should be brought against the defendants. Govt’s June 30, 2009 Letter at 4. Hulser testified that he had no recollection of being advised of this request. 22 Hr’g Tr., Oct. 23, 2009 p.m. at 43.

9. The First and Second Grand Juries

The Fifth Amendment guarantees that no civilian may be brought to trial ... ‘unless on a presentment or indictment of a Grand Jury.’ This constitutional guarantee presupposes an investigative body acting independently of either prosecuting attorney or judge, whose mission is to clear the innocent, no less than to bring to trial those who may be guilty. 23

A grand jury was convened in late November 2007, at which Frost, Murphy and Mealy testified, along with other members of the Raven 23 convoy. Govt’s Mem. at 10. Kohl and Ponticello were aware before these witnesses testified that media outlets had disseminated the defendants’ September 18 written statements in news reports and articles, and that many of the witnesses they intended to call, including Frost, Murphy and Mealy, had been exposed to the defendants’ written statements. Govt’s June 30, 2009 Letter at 5. Recognizing these concerns, the government advised the witnesses before they took the stand to testify based on what they had personally seen and heard, and not to disclose to the grand jury any information obtained from the sworn written statements of other Raven 23 team members. Id.

Prior to asking the grand jury to return the indictment, the trial team reviewed the grand jury record to determine whether any Garrity evidence had been presented. Id. The government concluded that despite the admonishments given to the witnesses, the grand jury may have been exposed to such evidence, specifically through the testimony of Frost, Murphy, Mealy and other *127 members of the Raven 23 convoy. Id,.; Hr’g Tr., Nov. 2, 2009 a.m. at 21-22. Consequently, the government elected to withdraw the case from this grand jury and to re-present the case to a second grand jury. Govt’s June 30, 2009 Letter at 5.

In preparation for the second grand jury, prosecutors Kohl and Malis met with Hulser and Karla Dobinski, a recognized DOJ Garrity expert, regarding the taint issue. Hr’g Tr., Nov. 2, 2009 a.m. at 8, 29. The prosecutors decided that, among the Blackwater witnesses, they would rely solely on the testimony of Frost, Murphy and Mealy, purportedly because the prosecutors were the most confident that these three witnesses could survive a Kastigar hearing. Id. at 24. Rather than have these witnesses take the stand again, however, the prosecutors chose to provide the second grand jury with transcripts of their testimony at the first grand jury, redacted to remove references to information derived from the defendants’ compelled statements. Id. at 31-32. Prosecutor Malis was responsible for making these redactions, in consultation with Hulser and Dobinski. Id. In addition, the prosecutors decided to have Robyn Powell, an FBI agent without any knowledge of or exposure to the case, summarize evidence for the second grand jury. Id. at 29-30.

The second grand jury was convened in late November 2008. Agent Powell, the sole live witness presented to the second grand jury, summarized evidence from Iraqi witnesses, as well as information from cooperating witness Ridgeway, who by this time had entered a guilty plea, and summarized portions of the prior grand jury testimony of Frost, Murphy and Mealy. See generally Govt’s Ex. 1. In addition, the prosecutors presented the second grand jury with redacted transcripts of the prior grand jury testimony of Frost, Murphy and Mealy, along with summaries prepared by the prosecutors of the evidence against each defendant. See generally id.

By proceeding in this fashion, the prosecutors withheld from the second grand jury substantial exculpatory evidence that had been presented to the first grand jury. For instance, Raven 23 team members Thomas Vargas, Jeremy Skinner, Daniel Childers and Edward Randall all testified before the first grand jury that the Raven 23 convoy responded to incoming fire. Vargas testified before the first grand jury that approximately “five seconds after we pulled into our positions, we started taking fire” and that he “could hear AK-47 fire” and “immediately saw two insurgents.” Hr’g Tr., Nov. 3, 2009 a.m. at 10. Skinner likewise testified that he heard gunfire and saw “two distinct separate muzzle flashes” fired by insurgents at the Raven 23 convoy. Id. at 12. Childers testified that he heard incoming gunfire coming from his seven to eight o’clock position. Id. at 17. And Randall testified that the Raven 23 convoy took fire from the south and southwest and that he saw a round impact the side of one of the vehicles. Id. at 18-19. Although Malis acknowledged that this testimony corroborated the defendants’ self-defense theory, none of this testimony was presented to the second grand jury. Id. at 10-19. Indeed, Malis testified that he chose not to present the testimony of these witnesses to the second grand jury because the testimony indicated that the witnesses were “hostile” to the prosecution. Hr’g Tr., Nov. 2, 2009 a.m. at 22. DOJ guidelines require prosecutors to present exculpatory evidence to the grand jury. United States Attorneys’ Manual § 9-11.233 (stating that “[i]t is the policy of the Department of Justice ... that when a prosecutor conducting a grand jury inquiry is personally aware of substantial evidence that directly negates the guilt of a subject of the investigation, the prosecu *128 tor must present or otherwise disclose such evidence to the grand jury before seeking an indictment against such a person”).

The government also redacted exculpatory portions of the testimony of Frost, Murphy and Mealy. For instance, Murphy testified at the first grand jury that even though he was positioned close to defendant Ball, he never saw Ball fire his weapon and that Ball certainly “wasn’t just shooting wildly” if he was shooting at all. Hr’g Tr., Nov. 2, 2009 p.m. at 115-16. The prosecutors redacted this testimony from the transcripts presented to the second grand jury, even though Malis acknowledged that it was exculpatory and had nothing whatsoever to do with the taint issue. Id. at 116.

Furthermore, the evidence summaries presented to the second grand jury presented distorted versions of the testimony on which they were based. For example, in the summary of evidence against defendant Slatten, the prosecutors stated that “Matt Murphy and Adam Frost recall that the first thing they heard at Nisur Square was the sound of two shots coming from the command vehicle, which they recognized as Slatten’s ‘heavy caliber’ sniper rifle.” GJ Ex. 107. 24 Agent Powell told the second grand jury that “[ajccording to Frost, the initial rounds that were fi[r]ed according to him sounded like they were from Slatten’s sniper rifle.” Hr’g Tr., Oct. 21, 2009 p.m. at 25. In reality, Frost had testified at the first grand jury that it was his “pure speculation” that those “two pops” were rifle fire. 25 Id. at 23-24. At the Kastigar hearing, Frost testified that Agent Powell’s “summary” of his testimony was inaccurate. Id. at 25.

Finally, Kohl informed the second grand jury at the outset of the proceedings that each of the defendants had given sworn testimony to State Department investigators in the immediate aftermath of the incident, but that “in exchange for them giving those statements, the U.S. government promised, essentially ... that [it] wouldn’t use those statements against them in a criminal proceeding.” Hr’g Tr., Nov. 3, 2009 a.m. at 22. This prompts the question — why did the prosecutors go out of their way to tell the grand jury that the defendants had made statements in exchange for immunity? The government’s explanation — that the prosecutors were attempting to determine if any grand jurors had been tainted by the defendants’ statements, Govt’s Reply Mem. at 17 — is unpersuasive, to say the least, as the prosecutors obviously could have inquired into whether any grand jurors had been exposed to the defendants’ statements without revealing that the defendants had claimed immunity. It is reasonable to believe that the purpose of these disclosures was to color the grand jury’s thinking by telling it of the existence, if not the content, of withheld incriminating statements made by the targets of the prosecution.

On December 4, 2008, the second grand jury returned an indictment against the defendants, charging them with voluntary manslaughter and weapons violations based on the Nisur Square incident. See generally Indictment.

B. Procedural History

In early May 2009, the government filed a “Motion for a Garrity Hearing in Lieu of *129 a Pretrial Kastigar Hearing,” in which it argued that the defendants’ statements to DSS investigators should not be analyzed as immunized statements, which would require the court to hold a Kastigar hearing to permit the government to prove it made no use of those statements. See generally Govt’s Mot. for a Garrity Hr’g in Lieu of a Pretrial Kastigar Hr’g. On July 22, 2009, the court denied the government’s motion, concluding that there had been inadequate briefing on the legal and factual foundation for the application of Kastigar. Mem. Order (July 22, 2009) at 4. The court set aside dates for a Kastigar hearing and ordered the defendants to file a motion setting forth a firm foundation for their assertion that immunized testimony was improperly utilized in this case. Id.

Following extensive briefing, the court issued a sealed memorandum opinion on September 8, 2009, holding that any Garrity compelled statements made by the defendants were entitled to use and derivative use immunity, and that the defendants had laid a firm foundation that immunized testimony was used in the investigation and prosecution of this case. See generally Mem. Op. (Sept. 8, 2009); infra Part III.A.2. Accordingly, the court ordered that a Kastigar hearing be held on the dates set aside in the court’s prior order. 26

The hearing, which commenced on October 14, 2009 27 and concluded three weeks later, explored two principal issues: (1) whether the defendants’ September 16, 2007 interview statements to DSS investigators constituted compelled statements under Garrity and (2) whether the government had made any impermissible use of any Garrity compelled statements. The parties acknowledged that the defendants had the burden of persuasion on the first issue while the government had the burden of persuasion on the second.

Approximately three weeks after the hearing concluded, while the parties were still engaged in post-hearing briefing, the government moved for leave to dismiss the indictment against defendant Slatten without prejudice, conceding that the evidence which came to light at the Kastigar hearing did not satisfy its Kastigar burden against that defendant. Slatten subsequently cross-moved for dismissal of the claims against him with prejudice, based on the absence of sufficient untainted evidence to support a subsequent indictment as well as prosecutorial misconduct in obtaining the indictment. Two weeks later, defendant Ball also filed a motion to dismiss the indictment against him with prejudice, alleging prosecutorial misconduct. 28

III. ANALYSIS

A. Legal Framework

1. Legal Standard for the Application of Kastigar

In Kastigar v. United States, 406 U.S. 441, 92 S.Ct. 1653, 32 L.Ed.2d 212 (1972), the Supreme Court held that

*130 ‘Once a defendant demonstrates that he has testified, under a state grant of immunity, to matters related to the federal prosecution, the federal authorities have the burden of showing that their evidence is not tainted by establishing that they had an independent legitimate source for the disputed evidence.’

This burden of proof ... is not limited to negation of taint; rather, it imposes on the prosecution the affirmative duty to prove that the evidence it proposes to use is derived from a legitimate source wholly independent of the compelled testimony.

Id. at 460, 92 S.Ct. 1653 (quoting Murphy v. Waterfront Comm’n, 378 U.S. 52, 79 n. 18, 84 S.Ct. 1594, 12 L.Ed.2d 678 (1964)). As this Circuit has noted, “[w]hen the government proceeds to prosecute a previously immunized witness, it has ‘the heavy burden of proving that all of the evidence it proposes to use was derived from legitimate independent sources.’ ” United States v. North, 910 F.2d 843, 854 (D.C.Cir.1990) (“North I ”) (quoting Kastigar, 406 U.S. at 461-62, 92 S.Ct. 1653), reh’g denied, 920 F.2d 940 (D.C.Cir.1990) (“North II’).

“A trial court must normally hold a hearing (a ‘Kastigar hearing’) for the purpose of allowing the government to demonstrate that it obtained all of the evidence it proposes to use from sources independent of the compelled testimony.” Id. (citing United States v. Rinaldi, 808 F.2d 1579, 1584 (D.C.Cir.1987)). During the hearing, the government must show that for each witness on whom it relies, “no use whatsoever was made of any of the immunized testimony either by the witness or by the [prosecutor] in questioning the witness.” Id. at 872-73 (noting that this inquiry “must proceed witness-by-witness; if necessary, it will proceed line-by-line and item-by-item”) (emphasis added). 29 “Unless the District Court can make express findings that the government has carried this heavy burden as to the content of all of the testimony of each witness, that testimony cannot survive a Kastigar test.” Id. at 873.

In articulating the reach of Kastigar’s prohibition on the “use” of immunized testimony, the Circuit has distinguished between “evidentiary” and “nonevidentiary” use. See North I, 910 F.2d at 856-61. Evidentiary use of immunized testimony includes the direct presentation of immunized testimony to the grand or petit jury, as well as any derivative (or indirect) use of the immunized testimony. See United States v. Hubbell, 530 U.S. 27, 41, 120 S.Ct. 2037, 147 L.Ed.2d 24 (2000) (observing that the introduction of evidence derived from immunized sources “would surely be a prohibited ‘use’ of the immunized act of production”). Derivative evidentiary use includes the exposure of witnesses to immunized testimony “to refresh their memories, or otherwise to focus their thoughts, organize their testimony, or alter their prior or contemporaneous statements,” North I, 910 F.2d at 860; see also North II, 920 F.2d at 942 (noting that “Kastigar is ... violated whenever the prosecution puts on a witness whose testimony is shaped, directly or indirectly, by compelled testimony, regardless of how or by whom he was exposed to that compelled testimony”). Evidentiary use also includes the use of immunized testimony to obtain investigatory leads, see United States v. Hubbell, 167 F.3d 552, 585 (D.C.Cir.1999), *131 and to influence a witness to testify, see United States v. Hylton, 294 F.3d 130, 134 (D.C.Cir.2002).

Nonevidentiary use, by contrast, is that which does not culminate directly or indirectly in the presentation of evidence against the immunized person. North I, 910 F.2d at 857. Such use includes “assistance in focusing the investigation, deciding to initiate prosecution, refusing to plea bargain, interpreting evidence, planning cross-examination and otherwise generally planning trial strategy.” Id. Although Kastigar clearly prohibits the government from making any direct or indirect evidentiary use of immunized testimony, see Kastigar, 406 U.S. at 460, 92 S.Ct. 1653, it “does not expressly discuss the propriety of nonevidentiary use,” North I, 910 F.2d at 858. The Circuit declined to resolve the issue in North I, see id. at 860 (assuming without deciding that Kastigar bars nonevidentiary use of immunized testimony), and has yet to squarely address the issue, see United States v. Kilroy, 27 F.3d 679, 687 n. 7 (D.C.Cir.1994).

Nonetheless, the court derives significant guidance regarding the permissible bounds of nonevidentiary use in the Circuit’s discussion of the issue in North I. The Circuit began its analysis by rejecting the rule articulated in United States v. McDaniel, 482 F.2d 305 (8th Cir.1973), which held that any nonevidentiary use necessarily violates the Fifth Amendment. See North I, 910 F.2d at 858-59 (citing McDaniel, 482 F.2d at 311). The Circuit noted its agreement with the holdings of other circuits that “a prosecution is not foreclosed merely because the ‘immunized testimony might have tangentially influenced the prosecutor’s thought processes in preparing for the indictment and preparing for trial.’ ” Id. at 859 (quoting United States v. Mariani 851 F.2d 595, 600 (2d Cir.1988)); see also United States v. Serrano, 870 F.2d 1, 16 (1st Cir.1989) (concluding that the McDaniel approach “would in effect grant a defendant transactional immunity once it is shown that government attorneys or investigators involved in the prosecution were exposed to the immunized testimony”). At the same time, the Circuit stated that insofar as the authorities criticizing McDaniel “may be read as establishing a rule that Kastigar allows nonevidentiary use of compelled testimony under all circumstances, we find those cases troubling.” North I, 910 F.2d at 859-60; cf Kastigar, 406 U.S. at 458-59, 92 S.Ct. 1653 (observing that in order not to violate the Fifth Amendment privilege against self-incrimination, “immunity from use and derivative use [must] ‘leave[ ] the witness and the Federal Government in substantially the same position as if the witness had claimed his privilege’ in the absence of a grant of immunity”) (quoting Murphy, 378 U.S. at 79, 84 S.Ct. 1594). Ultimately, the Circuit concluded in North I that given the absence of significant prosecutorial exposure to the immunized testimony in that case, the Circuit was not required to resolve the issue because the prosecution “could not have made significant nonevidentiary use” of the testimony. North I, 910 F.2d at 860.

The Circuit’s analysis in North I suggests that although no Kastigar violation occurs when a prosecutor’s fleeting exposure to immunized testimony has a merely tangential influence on his or her thoughts about a case, a Kastigar violation may result when a prosecutor has had significant exposure to immunized testimony and makes significant nonevidentiary use of that testimony. See id. at 859-60; see also United States v. Hsia, 131 F.Supp.2d 195, 201-02 (D.D.C.2001) (holding that Kastigar and North prohibit nonevidentiary uses of immunized testimony, in the context of a case in which the government made no effort to insulate the prosecution *132 from the compelled testimony); cf. United States v. Daniels, 281 F.3d 168, 182 (5th Cir.2002) (observing that “[t]here may be some cases in which the exposure of a prosecution team to a defendant’s immunized testimony is so prejudicial that it requires disqualification of the entire prosecution team”); United States v. Barker, 542 F.2d 479, 484 n. 9 (8th Cir.1976) (observing that in McDaniel, access to immunized testimony “had been obtained at a very early stage in the investigatory process”); United States v. McGuire, 45 F.3d 1177, 1183 (8th Cir.1995) (holding that McDaniel is limited to the “unusual circumstances” present in that case).

The government’s burden at a Kastigar hearing, though “heavy,” is not insurmountable. The government must demonstrate by a preponderance of the evidence that it made no use of the defendants’ testimony, a burden that the government may satisfy by showing that witnesses were never exposed to immunized testimony or that the allegedly tainted testimony contains no evidence that was not “canned” (or memorialized by investigators) before such exposure occurred. See North I, 910 F.2d at 872 (cautioning that “the government has to meet its proof only by a preponderance of the evidence, but any failure to meet that standard must result in exclusion of the testimony”). In addition, if the government has used immunized testimony, “[dismissal of the indictment or vacation of the conviction is not necessary where the use is found to be harmless beyond a reasonable doubt.” United States v. Ponds, 454 F.3d 313, 328 (D.C.Cir.2006). If, however, “the government has in fact introduced trial evidence that fails the Kastigar analysis, then the defendant is entitled to a new trial. If the same is true as to grand jury evidence, then the indictment must be dismissed.” North I, 910 F.2d at 873.

2. Legal Standard for the Application of Kastigar to Gairity Statements

In Garrity v. New Jersey, 385 U.S. 493, 87 S.Ct. 616, 17 L.Ed.2d 562 (1967), the Supreme Court held that the Fifth Amendment privilege against self-incrimination barred the use of statements obtained from police officers under a threat of job loss in a subsequent criminal proceeding. Id. at 500, 87 S.Ct. 616 (holding that “[t]he option to lose [one’s] means of livelihood or to pay the penalty of self-incrimination is the antithesis of free choice to speak out or remain silent”). The parties in this case supplied extensive briefing on whether the protections of Kastigar applied to Garrity compelled statements.

As noted in this court’s September 8, 2009 memorandum opinion, see generally Mem. Op. (Sept. 8, 2009), although Garrity does not expressly invoke the protections of Kastigar, every circuit to have addressed the issue has held that “a government employee who has been threatened with an adverse employment action by her employer for failure to answer questions put to her by her employer receives immunity from the use of her statements or their fruits in subsequent criminal proceedings,” Sher v. Dep’t of Veterans Affairs, 488 F.3d 489, 501-02 & n. 12 (1st Cir.2007); In re Grand Jury Proceedings (Kinamon), 45 F.3d 343, 348 (9th Cir.1995) (holding that statements obtained from an employee who was required to answer questions under a threat of dismissal were subject to use and derivative use immunity); In re Grand Jury Subpoenas (Stover), 40 F.3d 1096, 1102-03 (10th Cir.1994) (observing that “ Garrity’s protection ... acts to immunize ... compelled statements, as it prohibits their subsequent use against the officer so as not offend the Fifth Amendment Privilege” and that this prohibition “provides a comprehensive safe *133 guard, barring the use of compelled testimony as an ‘investigatory lead,’ and also barring the use of any evidence obtained by focusing investigation on a witness as a result of his compelled disclosures”) (quoting Kastigar, 406 U.S. at 460, 92 S.Ct. 1653); In re Grand Jury Proceedings (Cohen), 975 F.2d 1488, 1490 (11th Cir.1992) (observing that “[ijmmunity under Garrity prevents any statements made in the course of the internal investigation from being used against the officers in subsequent criminal proceedings”); Uniformed Sanitation Men Ass’n v. Comm’r of Sanitation of the City of N.Y., 426 F.2d 619, 626 (2d Cir.1970) (stating that in Garrity, “the very act of the attorney general in telling the witness that he would be subject to removal if he refused to answer was held to have conferred ... immunity”).

Furthermore, courts have uniformly held that “if a criminal defendant ... demonstrates that she was compelled to testify by her government employer, ‘the government must show that any evidence used or derived has a legitimate source wholly independent of the compelled testimony.’ ” United States v. Moten, 551 F.3d 763, 766 (8th Cir.2008) (quoting Kastigar, 406 U.S. at 460, 92 S.Ct. 1653); In re Grand Jury, Doe, 478 F.3d 581, 583-84 (4th Cir.2007) (noting that if a police officer believed statements compelled under Garrity “were used to indict him, he would be entitled to a Kastigar hearing, at which the government would bear the burden of ‘proving] that the evidence it proposes to use is derived from a legitimate source wholly independent of the compelled testimony’ ”) (quoting Kastigar, 406 U.S. at 441, 92 S.Ct. 1653); Kinamon, 45 F.3d at 348 (ordering the district court to conduct a Kastigar hearing to determine whether statements obtained by threat of job loss were improperly used in a grand jury proceeding); Stover, 40 F.3d at 1103 (holding that “[i]f an officer, whose compelled statement [under Garrity] has been considered by the grand jury, ultimately is indicted, that officer will be able to challenge the indictment and the government will be required to prove that its evidence derives entirely from legitimate sources or that the grand jury’s exposure to the officer’s statement was harmless”). Thus, the court concluded that statements compelled under Garrity are entitled to the full panoply of protections that Kastigar provides to other immunized statements. See generally Mem. Op. (Sept. 8, 2009).

B. The Defendants’ September 16 Interview Statements Were Compelled

At the outset of the Kastigar hearing, the government conceded that the defendants’ September 18 written statements and all of the defendants’ subsequent statements to DSS investigators constituted compelled statements under Garrity. 30 See Hr’g Tr., Oct. 14, 2009 a.m. at 6-10. In light of the court’s September 8, 2009 ruling, the government acknowledged its obligation to demonstrate that it had made no use or derivative use of those statements. Id. The government has maintained, however, that any statements that the defendants made during their September 16, 2007 interviews were not compelled under Garrity and are therefore not entitled to any form of immunity. Id.; Govt’s Mem. at 17. This court disagrees.

To demonstrate compulsion under Garrity, a public employee must show (1) that he subjectively believed his statements were compelled by the threat of job *134 loss and (2) that this belief was objectively reasonable. See United States v. Friedrich, 842 F.2d 382, 395 (D.C.Cir.1988); accord United States v. Vangates, 287 F.3d 1315, 1321-22 (11th Cir.2002). The government does not dispute that the defendants subjectively believed that they would have been terminated had they refused to submit to questioning from DSS agents on September 16, 2007. 31 Govt’s Mem. at 18. Rather, the government contends that the defendants’ belief was not objectively reasonable because on September 16, the DSS agents were merely documenting (and not investigating) the Nisur Square incident. Id. at 23-26. In a similar vein, the government argues that the September 16 statements resulted from routine, job-related reporting requirements that do not implicate Garrity. Id. at 18-23. The defendants point out that ample evidence supports the reasonable objectiveness of their belief that refusing to submit to the September 16, 2007 interviews would have meant termination, and assert that that the September 16 interviews were anything but routine. See Defs.’ Mem. at 5-7; Defs.’ Post-Hr’g Reply Mem. (“Defs.’ Reply Mem.”) at 3-6.

1. The Objective Reasonableness of the Defendants’ Belief That Their September 16 Interview Statements Were Compelled Under the Totality of the Circumstances

When an employee has been expressly advised by his employer that he must submit to questioning or face termination, his belief that he will be terminated for refusing to comply is objectively reasonable. See Vangates, 287 F.3d at 1321 (observing that “Garrity is more easily applied to situations ‘[wjhere the state has directly presented the defendant with the Hobson’s choice of either making an incriminating statement or being fired’ than to cases ... in which there has been no direct threat of termination”) (citing United States v. Camacho, 739 F.Supp. 1504, 1515 (S.D.Fla.1990)). This Circuit has made clear, however, that the absence of an express Garrity warning does not indicate a lack of objective reasonableness. See Friedrich, 842 F.2d at 395-96. The defendant in Friedrich was an FBI agent under investigation for suspected misconduct. Id. at 384. Between June and September 1985, he was subjected to a series of interviews, during which he was provided with an FBI Form 645, which advised the agent that his refusal to provide information could result in dismissal, but that the government could not use any information provided against the agent in a subsequent criminal prosecution. Id. at 385-87. In January 1986, the FBI again interviewed the defendant, this time without presenting him with a Form 645 or mentioning immunity in any way. Id. at 387. The district court held that the statements made by the defendant during the January 1986 interviews were inadmissible, despite the absence of express Garrity warnings, based on the Fifth Amendment principles articulated in Garrity and Kastigar. Id. at 395-96.

The Circuit affirmed, concluding that under the totality of circumstances, the defendant reasonably believed that the same rules that had applied during the interviews in the fall of 1985 also governed the January 1986 interviews. Id. at 396-97. The Circuit noted that the defendant’s supervisors directed him to appear in Washington D.C. for the January 1986 interviews in precisely the same manner in *135 which he had been ordered to appear for his previous interviews. Id. The Circuit also noted that the defendant had been ordered to appear for “further interviews,” indicating that the January 1986 interviews were a continuation of the interviews conducted in the fall of 1985. Id. The Circuit considered the absence of a Form 645 at the January 1986 interviews to be of little moment, as the defendant reasonably would have expected the interviewing agents to advise him if the nature of the interviews had changed, given the pervasive practice of providing him with Form 645 warnings during prior interviews. Id. (stating that “studied silence in this context is consistent with a continuation of the already established pattern; it is emphatically not a signal of an investigative volte face ”). In summary, the Circuit explained that

[t]he reason for our conclusion is simple and straightforward: Friedrick was at all times during the questioning an FBI employee who was under orders from his superiors in the Bureau to attend and participate in the interviews. Friedrick had not been invited to come to Washington. He was instructed to appear, just as he had been several times previously. Friedrick was plainly in Washington ... in the capacity of an FBI employee who had no choice but to answer questions or else risk being fired.

Id. at 398.

During the hearing held in this case, the court was presented with substantial evidence indicating that it was objectively reasonable for the defendants to believe that that they would be fired if they refused to answer questions on September 16, 2007. Agent Carpenter testified that he directed all the Raven 23 team members to come to the Palace for questioning, and that he did not consider this instruction to be voluntary. Hr’g Tr., Oct. 19, 2009 a.m. at 99-100. Moreover, every Raven 23 team member who appeared at the hearing — including Mealy, Murphy and Frost, who testified on behalf of the prosecution — testified that they were directed to submit to questioning on September 16, 2007 and that they understood that their refusal to answer the DSS agents’ questions would have resulted in termination. Mealy testified that although he had no specific recollection, he “must have been directed” to submit to questioning on September 16, 2007. Id. at 44-45. He further testified that he understood that he “very well could have been fired” had he refused to answer the DSS agents’ questions on that date. Id. Murphy testified that that he was directed to appear for his September 16, 2007 interview and that it was “a given” that his refusal to submit to questioning that day would result in his termination. Hr’g Tr., Oct. 14, 2009 p.m. at 22.

Frost submitted an affidavit in which he stated that

[o]n September 16, 2007, I was ordered, along with the other members of Raven 23, by Blackwater management to report to the United States Embassy in Baghdad to provide a verbal statement to Department of State investigators describing my actions during the Nisour Square incident. I understood this order to be mandatory — that being interviewed was required by the Department of State, and was a condition of my employment. I understood then, and believe now, that had I refused the order I would have been terminated.

Frost Ex. 1 ¶ 5. 32 At the Kastigar hearing, Frost testified that he still believed this *136 portion of his declaration to be accurate and that he understood, at the time he gave his September 16 interview statement to the DSS investigators, that he was required to do so. Hr’g Tr., Oct. 21, 2009 p.m. at 6-9.

Daniel Childers, another Raven 23 team member, testified that he was ordered to appear for an interview on September 16, 2007, that he did not believe that he had a choice in the matter and that he understood at the time that if he had refused to submit to questioning, he would have been fired. Hr’g Tr., Oct. 29, 2009 p.m. at 19-21. And although he did not testify at the Kastigar hearing, the government has conceded that cooperating witness Ridgeway “does subjectively believe that the members of Raven 23 could have had their contracts terminated if they did not provide statements to DSS.” Govt’s June 30, 2009 Letter at 9. The fact that every testifying Raven 23 member, including all those supporting the government’s prosecution, corroborated the defendants’ subjective belief that they would have been terminated had they refused to submit to questioning strongly indicates that their subjective belief was objectively reasonable.

This uniformly held belief among Raven 23 likely resulted from their prior experiences with the reporting procedures outlined in the Hunter Memorandum. The Hunter Memorandum, which all Raven 23 members received when they arrived in Iraq, required Blackwater personnel to report any incident of firearms discharge to State Department officials, to submit to oral debriefing following the incident and to provide a sworn written statement on a form attached to the memorandum. Govt’s Ex. 32 at 1-3. That form contained the following express Garrity warning:

I understand that this statement is made in furtherance of an official administrative inquiry regarding potential misconduct or improper performance of official duties and that disciplinary action, including dismissal from the Department’s Worldwide Personnel Protective Services contract, may be undertaken if I refuse to provide this statement or fail to do so fully and truthfully. I further understand that neither my statements nor any information or evidence gained by reason of my statements can be used against me in a criminal proceeding.

Id. at 3. Thus, the Hunter Memorandum expressly required Blackwater personnel to provide sworn written statements to State Department officials following any incident of firearms discharge or face termination. Id. The Hunter Memorandum did not specify that the threat of job loss applied only to the obligation to provide a sworn written statement to the State Department and not to any prior or subsequent statements requested by the State Department investigators regarding that incident. See generally id.

Prior to September 16, 2007, all of the defendants had been involved in shooting incidents, after which they had been required to provide sworn written statements under the express Garrity warning on the State Department form. See Hr’g Tr., Oct. 27, 2009 a.m. at 10-11; Hr’g Tr., Nov. 3, 2009 p.m. at 49-50, 82-83, 117, 136-37. During the September 16, 2007 interviews, the DSS agents did not advise the Raven 23 team members that the interviews were voluntary or that they were not required to answer questions. Hr’g Tr., Oct. 15, 2009 a.m. at 75; Hr’g Tr., Oct. 15, 2009 p.m. at 40-41,121. In fact, Agent Scollan testified that under the rules then *137 in place, 33 the defendants should have been provided Garrity warnings during the September 16, 2007 interviews. Hr’g Tr., Oct. 15, 2009 a.m. at 86-87. Under these circumstances, and in light of their prior experiences, it was reasonable for the defendants (and the other members of Raven 23) to believe that their September 16 interview statements, like statements they had provided after other shooting incidents, were compelled under threat of job loss.

Furthermore, although the interviewing DSS agents testified that they did not warn the defendants that they would be fired if they refused to answer the agents’ questions, id. at 8; Hr’g Tr., Oct. 15, 2009 p.m. at 25, three of the defendants testified that the interviewing DSS agents told them that their statements could not be used against them in a criminal proceeding, see Hr’g Tr., Nov. 3, 2009 p.m. at 52, 116-17, 137-38. These admonitions were consistent with the Garrity warning contained on the State Department form, and reasonably could have contributed to a belief by those defendants that the same conditions applied to statements made during the September 16, 2007 interviews as well.

In sum, the facts adduced at the Kastigar hearing bring this case squarely within the bounds of Friedrich. As in Friedrich, the defendants in this case were directed to submit to questioning. As in Friedrich, the defendants had provided sworn statements on prior occasions under express Garrity warnings. And as in Friedrich, the defendants were not advised that the ground rules had changed — more specifically, that the rules that had governed the defendants’ other post-incident statements to the State Department did not apply on September 16, 2007. Furthermore, three of the defendants in this case testified that they were expressly told during their September 16, 2007 interviews that their statements could not be used against them in a criminal proceeding. Moreover, the defendants elicited corroborating testimony from friendly and hostile witnesses alike, all of whom testified that they too believed that they were required to submit to questioning on September 16, 2007 under threat of job loss. The court is persuaded that this evidence establishes the objective reasonableness of the defendants’ belief that they were required to submit to questioning or face job loss.

2. The Government’s Reliance on United States v. Cook and Related Authorities

The government contends that the September 16 interview statements do not implicate Garrity because they resulted from mundane, job-related reporting requirements, not unlike the obligation to report a missing or stolen weapon. See Govt’s Mem. at 19-23. Indeed, routine, contemporaneously-prepared incident reports are generally not considered compelled Garrity statements. See, e.g., United States v. Cook, 526 F.Supp.2d 1, 8-9 (D.D.C.2007) (rejecting the defendant’s Garrity challenge to a use-of-force report that he was ordered to prepare after an individual in his custody complained that the defendant had assaulted him), affd 330 Fed.Appx. 1 (D.C.Cir.2009); see also Devine v. Good- *138 stein, 680 F.2d 243, 246-47 (D.C.Cir.1982); United States v. Ruiz, 579 F.2d 670, 675-76 (1st Cir.1978); Watson v. County of Riverside, 976 F.Supp. 951, 954-55 (C.D.Cal.1997). Yet the government’s reliance on Cook and related authorities is misplaced, as those cases are distinguishable from this matter in at least two critical respects.

a. The September 16, 2007 Interviews Were Far From Routine

First, Cook and the related authorities cited by the government rest on the principle that the element of compulsion is absent when a report is so routine that a defendant could not have reasonably believed that the report could form the basis of a criminal prosecution. See Cook, 526 F.Supp.2d at 8 (observing that “Garrity does not stand for the proposition that a statement made in a standard report is coerced whenever an officer faces both the remote possibility of criminal prosecution if he files the report and the arguably even more speculative possibility of termination if he declines to do so”); 34 see also Devine, 680 F.2d at 247 (observing that Garrity only protects against “disclosures that the witness reasonably believes could be used in a criminal prosecution or could lead to other evidence that might be so used”); Watson, 976 F.Supp. at 955 (noting that the element of compulsion was absent because the creation of the arrest report was simply a product of the defendant’s normal job requirements).

The September 16, 2007 interviews, however, were anything but routine. Most obviously, the interviews were conducted in response to an extraordinarily violent shooting incident that resulted in the death or injury of more than thirty people. See Govt’s Mem. at 2. As just one example, Agent Lopez testified that she had never before encountered anything like the Nisur Square incident and that she did not view it as a routine situation. Hr’g Tr., Oct. 15, 2009 p.m. at 39-40.

Furthermore, by all accounts, it was out of the ordinary for DSS agents to conduct interviews of all the members of a unit following a shooting incident. Agent Carpenter testified that he had never before directed every member of a detail to submit to questioning as he did on September 16, 2007, and that he was unaware of any prior instance in which that procedure had been followed. Hr’g Tr., Oct. 19, 2009 a.m. at 99-100. David Farrington, another DSS agent who assisted in the DSS investigation, testified that he was unaware of any incident in which every member of a unit was interviewed following a shooting incident. Hr’g Tr., Oct. 19, 2009 p.m. at 57-58. Raven 23 team member Childers also testified that it was unusual for the entire team to be asked to report for oral interviews. Hr’g Tr., Oct. 23, 2009 a.m. at 20.

In fact, there was considerable doubt as to whether it was, in practice, standard procedure for oral interviews to be conducted after shooting incidents at all. Defendant Slough testified that although he had been involved in two prior shooting incidents in Iraq, September 16, 2007 was the first time he had been required to give an oral statement to DSS agents and the first time his entire unit had been required to submit to questioning following a shooting incident. Hr’g Tr., Nov. 3, 2009 p.m. at 49-51. Likewise, defendant Ball testified that although he had provided written statements following incidents involving the discharge of a weapon on at least two prior occasions, he had never been interviewed by a State Department agent. Id. at 135-36. Prosecutor Ponticello testified *139 that the September 16, 2007 interviews were “the first time we heard that oral interviews were given,” that “it was either Mr. Frost or Mr. Mealy [who] may have testified that it was unusual for that to happen,” Hr’g Tr., Oct. 26, 2009 p.m. at 72-73, and that based on his understanding, the process of having the Raven 23 members go to the U.S. Embassy for questioning was not routine, Hr’g Tr., Oct. 27, 2009 p.m. at 57-58. Agent Carpenter testified that when he reviewed the case files of prior shooting incidents, he found that they “consisted mainly of the statements of those involved that had fired their weapon, an after-action report by the team leader, and maybe a statement from another teammate in the detail [who] ... was there at the incident.” 35 Hr’g Tr., Oct. 16, 2009 p.m. at 105. This evidence indicates that the defendants’ September 16 oral interview statements were not the product of routine, after-incident reporting, like the statements at issue in Cook.

b. The September 16, 2007 Were Conducted as Part of the DSS Investigation

The second distinguishing feature of Cook and the other cases cited by the government is that they concerned statements generated prior to the initiation of any investigation into the defendant’s conduct. See Cook, 526 F.Supp.2d at 8 (observing that “the presumption underlying Garrity and its progeny is that the subject employee is under investigation at the time the challenged statement is made”); see also Devine, 680 F.2d at 247; Ruiz, 579 F.2d at 675-76; Watson, 976 F.Supp. at 954-55. 36 In this case, however, there is ample evidence that the September 16, 2007 interviews were conducted as part of an investigation into the defendants’ actions at Nisur Square. As an initial matter, the government admits that the shooting was extraordinary in its violence and points out that in its immediate aftermath, U.S. and Iraqi forces flooded the traffic circle and began documenting the scene and collecting physical evidence. Govt’s Mem. at 2, 7-8. Two experienced, high-ranking U.S. military officers, Colonels Boslego and Tarsa, arrived on the scene less than an hour after the incident and reportedly found little evidence of insurgent activity. Id. Accordingly, the specter of improper and potentially criminal conduct was apparent to government officials almost immediately after the incident.

*140 Agent Carpenter testified that from September 16, 2007 onward, he and his agents “were conducting an investigation. We were trying to get to what happened that day. My initial indications [were] that ... Blackwater personnel were responding to some type of ambush, and I was trying to get to that issue.” Hr’g Tr., Oct. 19, 2009 p.m. at 103-04. Accordingly, Agent Carpenter acknowledged in his investigative report that “[o]n September 16, 2007, the Regional Security Office (RSO) initiated an administrative investigation into the application of deadly force by Personal Protective Specialists (PSS) assigned to [Raven 23].” Defs.’ Carpenter Ex. 6(a) at 1.

The fact that the DSS agents were not merely “documenting” the shooting on September 16, 2007 is further illustrated by current State Department investigative guidelines. These guidelines provide for an initial debriefing following any serious incident “for the purpose of gathering information needed to assure appropriate incident response and management, and to assure that the operational responsibilities of the RSO related to the security of mission personnel are properly carried out”— in other words, expressly the type of routine initial debriefing that the government argues was taking place on September 16, 2007. Govt’s Ex. 35 at 4. Yet these guidelines require only “the Shift Leader” and the “Tactical Commander,” rather than every member of the detail, to report for this initial debriefing. Id. This procedure, of course, is consistent with the purpose of routine State Department debriefing, which is “to learn quickly the key and readily-available facts of an incident (who, what, when, where).” Id. The debriefing is expressly “not intended to cover all known details.” Id.

Agent Carpenter testified that immediately after the Raven 23 convoy returned to the U.S. Embassy, Jim Watson and Cory Wainscott, Raven 23’s Shift Leader and Tactical Commander, see Govt’s Ex. 2, provided an initial debriefing to summarize the events that had occurred at Nisur Square. Hr’g Tr., Oct. 16, 2009 p.m. at 115-16. Although Agent Carpenter acknowledged that this debriefing provided him with a “basic understanding” of what had occurred, he nonetheless ordered all nineteen members of Raven 23 to appear for interviews to obtain additional details about the incident. Id. at 116.

More fundamentally, it is beyond dispute that at some point, Agent Carpenter and his DSS agents were conducting a full blown investigation into the defendants’ actions at Nisur Square. Hr’g Tr., Oct. 19, 2009 p.m. at 103-04. In the days following the shooting, these agents repeatedly interviewed those Raven 23 members who had acknowledged firing their weapons to obtain details of their accounts. See Defs.’ Scollan Exs. 3-5; Defs.’ Lopez Ex. 5. They also conducted a physical search of Nisur Square to test the shooters’ accounts. Defs.’ Carpenter Ex. 8. The court sees no persuasive justification for treating the defendants’ September 16 interview statements — which were taken by the same agents who conducted the subsequent investigation and concerned the same events described in subsequent compelled statements — as wholly distinct from the investigation that followed. Rather, the court is persuaded that the September 16, 2007 interviews comprised the initial stage of an administrative investigation into a uniquely violent event, an investigation that continued over several days and culminated in the preparation of the investigative report prepared by Agent Carpenter and provided to the Justice Department.

In light of the evidence described above, the court concludes that the September 16, 2007 interviews were not conducted pursu *141 ant to any routine, standardized reporting requirement, but were instead taken as part of the DSS investigation into Raven 23’s actions at Nisur Square. Accordingly, the government’s reliance on Cook and similar authorities is misplaced.

3. The Allegedly False and Exculpatory Nature of the Accounts

Lastly, the government argues that the defendants’ September 16 interview statements (as well as their subsequent oral and written statements to DSS investigators) are not entitled to Fifth Amendment protection because they contain nothing more than the defendants’ false, exculpatory accounts of the events at Nisur Square. Govt’s Mem. at 25, 44 n. 203. The defendants respond that the content of the defendants’ statements has no bearing on whether Garrity applies. Defs.’ Mem. at 5-6; Defs.’ Reply Mem. at 22-24.

The government’s assertion that the falsity of the defendants’ statements renders inapplicable the Fifth Amendment privilege against self-incrimination requires little discussion. The Supreme Court has held that the question of coercion under the self-incrimination clause “is to be answered with complete disregard of whether or not [the accused] in fact spoke the truth.” Rogers v. Richmond, 365 U.S. 534, 544, 81 S.Ct. 735, 5 L.Ed.2d 760 (1961); Shotwell Mfg. Co. v. United States, 371 U.S. 341, 351 & n. 10, 83 S.Ct. 448, 9 L.Ed.2d 357 (1963) (concluding that although the self-incrimination clause did not prohibit prosecutorial use of false statements made pursuant to a voluntary disclosure policy, “[a] quite different case would be presented if an offer of immunity had been specifically directed to petitioners in the context of an investigation, accusation, or prosecution” because “the truth or falsity of such a disclosure would then be irrelevant to the question of its admissibility”); North I, 910 F.2d at 861 (observing that “Kastigar addresses ‘use,’ not ‘truth’ ”). 37

The authorities cited by the government indicate only that the Fifth Amendment does not prohibit the government from prosecuting a defendant for making a false statement under a grant of immunity. See United States v. Apfelbaum, 445 U.S. 115, 121-22, 100 S.Ct. 948, 63 L.Ed.2d 250 (1980) (observing that the federal immunity statute “makes no distinction between truthful and untruthful statements made during the course of the immunized testimony” but instead “creates a blanket exception from the bar against the use of immunized testimony in cases in which the witness is subsequently prosecuted for making false statements”); cf. Brogan v. United States, 522 U.S. 398, 404, 118 S.Ct. 805, 139 L.Ed.2d 830 (1998) (affirming the defendant’s conviction for making a false statement because “neither the text nor the spirit of the Fifth Amendment confers a privilege to lie”); United States v. Knox, 396 U.S. 77, 82, 90 S.Ct. 363, 24 L.Ed.2d 275 (1969) (holding that the Fifth Amendment did not prohibit the use of a compelled statement in a prosecution for submitting false information on a federal *142 form); United States v. White, 887 F.2d 267, 273-74 (D.C.Cir.1989) (affirming the defendant’s conviction for making a false statement); United States v. Veal, 153 F.3d 1233, 1240 (11th Cir.1998) (holding that Garrity does not prohibit a subsequent perjury or obstruction of justice charge if a witness makes a false statement); Cook, 526 F.Supp.2d at 9 (noting that even if the defendant could demonstrate that Garrity applied to statements made in a written reports, Garrity would not prevent the introduction of his reports as evidence in connection with the false statements and obstruction of justice charges levied against him). Accordingly, the court concludes that the alleged falsity of certain portions of the defendants’ statements to DSS investigators has no bearing on whether these statements are compelled for purposes of Garrity and Kastigar.

As for the exculpatory nature of the defendants’ accounts, the Supreme Court has stated that “[t]he Fifth Amendment prohibits only compelled testimony that is incriminating.” Hiibel v. Sixth Judicial Dist. Court of Nev., Humboldt County, 542 U.S. 177, 189, 124 S.Ct. 2451, 159 L.Ed.2d 292 (2004) (emphasis added). Yet the Court has defined “incrimination” broadly to encompass not only “answers that would in themselves support a conviction ... but likewise ... those which would furnish a link in the chain of evidence needed to prosecute the [defendant].” Hoffman v. United States, 341 U.S. 479, 486, 71 S.Ct. 814, 95 L.Ed. 1118 (1951). “Compelled testimony that communicates information that may ‘lead to incriminating evidence’ is privileged even if the information itself is not inculpatory.” Hubbell, 530 U.S. at 38, 120 S.Ct. 2037 (quoting Doe v. United States, 487 U.S. 201, 208 n. 6, 108 S.Ct. 2341, 101 L.Ed.2d 184 (1988)); Kastigar, 406 U.S. at 445, 92 S.Ct. 1653 (remarking that the Fifth Amendment privilege against self-incrimination “protects against any disclosures that the witness reasonably believes could be used in a criminal prosecution or could lead to other evidence that might be so used ”) (emphasis added).

Indeed, the Supreme Court has made clear that it has “never held that the privilege is unavailable to those who claim innocence. To the contrary, [the Court has] emphasized that one of the Fifth Amendment’s ‘basic functions ... is to protect innocent men ... who otherwise might be ensnared by ambiguous circumstances.’ ” Ohio v. Reiner, 532 U.S. 17, 21, 121 S.Ct. 1252, 149 L.Ed.2d 158 (2001) (reversing the Supreme Court of Ohio’s holding that a witness’s assertion of innocence deprived her of her Fifth Amendment privilege against self-incrimination) (quoting Grunewald v. United States, 353 U.S. 391, 421, 77 S.Ct. 963, 1 L.Ed.2d 931 (1957) (internal quotation marks omitted)). The defendant in Reiner, who had been convicted of involuntary manslaughter in connection with the death of his son, claimed that his son’s babysitter was the individual responsible for the death. Id. at 18, 121 S.Ct. 1252. Because the babysitter indicated that she would assert her Fifth Amendment privilege if called to testify, the trial court granted the government’s request that she be given transactional immunity for her testimony. Id. At trial, she testified that she had had no involvement in the child’s death and had not caused any of his injuries. Id. at 18-19, 121 S.Ct. 1252. The Supreme Court of Ohio held that the trial court’s grant of immunity was unlawful because the fact that the witness asserted her innocence made it impossible for her to invoke a Fifth Amendment privilege against self-incrimination. Id. at 19, 121 S.Ct. 1252. The Supreme Court reversed, noting that “[i]t need only be evident from the implications of the question, in the setting in which it is asked, that a respon *143 sive answer to the question or an explanation of why it cannot be answered might be dangerous because injurious disclosure could result.” Id. (quoting Hoffman, 341 U.S. at 486, 71 S.Ct. 814). The Court concluded that the babysitter had reasonable cause to believe that her answers could expose her to risk if questioned at trial, given that she had spent extended periods of time alone with the victim immediately preceding his injuries, she was with the victim within the potential time-frame of the fatal trauma and the defense’s theory was that she was responsible for the child’s death. Id. at 22, 121 S.Ct. 1252.

For the reasons previously discussed, see supra Part III.B.1, the court concludes that it was evident to the defendants, based on the circumstances under which they made their statements to DSS investigators, that their answers could have resulted in injurious disclosure. More specifically, these statements, given to government investigators immediately after an extraordinarily violent incident in which they had all played some role, disclosed a host of information potentially injurious to each of the defendants, including the fact that he fired his weapons, the type of weapons he used and the directions in which he fired. Even the statements made by defendant Liberty, who did not acknowledge firing his weapon, revealed particular details of the insurgent activity he allegedly saw and the specific locations of the threats he allegedly perceived. There is little doubt that some or all of this information could form a link in the chain of evidence used against the defendants, and at the very least, could have led to the discovery of inculpatory evidence. See Hubbell, 530 U.S. at 38, 120 S.Ct. 2037; Hoffman, 341 U.S. at 486, 71 S.Ct. 814.

In each of the authorities on which the government relies, the court’s determination that there was no Kastigar violation turned not on the fact that the immunized statement was exculpatory, but rather on the fact that the prosecution did not use the defendant’s statement against him or her. See United States v. Bartel, 19 F.3d 1105, 1113 (6th Cir.1994) (holding that the government had met its Kastigar burden and that the defendant’s “exculpatory, non-incriminating testimony did not contribute in any conceivable manner to his being indicted”); United States v. Caporale, 806 F.2d 1487, 1518 (11th Cir.1986) (noting that the immunized testimony “was self-serving and of no real value in the subsequent investigation”); United States v. Anderson, 450 A.2d 446, 451 (D.C.1982) (finding that prosecutors did not use the defendant’s exculpatory account to obtain investigatory leads or to focus the investigation). The government has cited to no authority indicating that the exculpatory nature of a defendant’s statement automatically renders the Fifth Amendment privilege against self-incrimination inapplicable, and indeed, such a holding would be inconsistent with the Supreme Court’s reasoning in Reiner. See Reiner, 532 U.S. at 21, 121 S.Ct. 1252; cf. Illinois v. Perkins, 496 U.S. 292, 303, 110 S.Ct. 2394, 110 L.Ed.2d 243 (1990) (noting that the government “may not use statements, whether exculpatory or inculpatory, stemming from custodial interrogation of the defendant unless it demonstrates the use of procedural safeguards effective to secure the privilege against self-incrimination”) (quoting Miranda v. Arizona, 384 U.S. 436, 444, 86 S.Ct. 1602, 16 L.Ed.2d 694 (1966)). Accordingly, the mere fact that the defendants maintained their innocence in the statements given to the DSS investigators does not preclude them from invoking the Fifth Amendment protections articulated in Garrity and Kastigar.

*144 C. The Government Failed to Prove by a Preponderance of the Evidence That It Made No Impermissible Use of the Defendants’ Compelled Garrity Statements

The defendants maintain that the government made impermissible evidentiary and nonevidentiary use of the defendants’ compelled Garrity statements in prosecuting this case. Noting that the prosecutors, investigators and government witnesses were thoroughly exposed to the defendants’ compelled statements, they contend that the taint of the defendants’ compelled statements has infiltrated nearly every aspect of this prosecution and that as a result, the indictment must be dismissed. The government asserts, however, that it made no impermissible use of the defendants’ compelled statements and that any such use that may have occurred was harmless beyond a reasonable doubt. The court now turns to the specific categories of potential “use” that form the heart of this dispute.

1. Evidentiary Use of the Defendants’ Garrity Statements

a. The Statements of Frost and Murphy Before the First and Second Grand Jury Were Tainted

Without question, the testimony of Frost, Murphy and Mealy played a critical role in the government’s presentations to the first and second grand juries. These witnesses identified which members of the Raven 23 convoy were the shooters, provided detailed accounts of the events that occurred at Nisur Square and testified regarding the absence of any threat justifying the defendants’ actions. See generally GJ Exs. 89-94. Among all of the Raven 23 team members who testified at the first grand jury, the prosecutors chose to rely solely on the testimony of Frost, Murphy and Mealy to obtain the indictment from the second grand jury. See generally id.

It is equally clear that by the time they testified at the first grand jury, Frost and Murphy had been thoroughly immersed in the defendants’ compelled Garrity statements. 38 Media reports alluding to the statements made to DSS investigators began to appear almost immediately after the incident, see, e.g., Defs.’ Media Exs. 23, 28, and excerpts from the defendants’ written statements began to appear shortly thereafter, see, e.g., Defs.’ Media Exs. 2, 6, 7, 13, 14, 23. Frost indicated that in the days following the shooting, he read numerous news reports on the Internet regarding the events at Nisur Square. Defs.’ Frost Ex. 6 at 1. Frost and Murphy both set up “Google News” alerts, through which they were e-mailed multiple news articles each day regarding the Nisur Square incident. Hr’g Tr., Oct. 21, 2009 p.m. at 26-27; Hr’g Tr., Oct. 14, 2009 p.m. at 27. Frost and Murphy both acknowledged reading the September 18 written statements of defendants Slatten and Slough, and Murphy further acknowledged reading the written statement of defendant Ball as well, all of which had been posted on the Internet. Hr’g Tr., Oct. 21, 2009 p.m. at 25-26; Hr’g Tr., Oct. 14, 2009 p.m. at 6, 74-75, 88-89. And on October 2, 2007, Frost and Murphy received an e-mail from Raven 23 team member Edward Randall containing a link to an ABC News televised report, which purported to be based on the written statements of all nineteen Raven 23 members and quoted extensively from several statements, including Slough’s. See Defs.’ Murphy Ex. 8; Hr’g Tr., Oct. 14, 2009 p.m. at 7. Nei *145 ther Frost nor Murphy recalled ever being warned by the government’s trial team to avoid exposure to news coverage about the incident or to the sworn statements given by the defendants. Hr’g Tr., Oct. 14, 2009 p.m. at 64; Hr’g Tr., Oct. 21, 2009 p.m. at 27.

At the Kastigar hearing, both Frost and Murphy testified that despite their unbounded exposure to the defendants’ compelled statements, they could testify based on their personal recollections alone and segregate any information learned from the compelled statements they had read. Hr’g Tr., Oct. 14, 2009 p.m. at 16; Hr’g Tr., Oct. 21, 2009 a.m. at 97. Yet when government prosecutors reviewed the testimony given by Frost and Murphy to the first grand jury, they saw signs that numerous portions of the testimony were influenced by the witnesses’ exposure to the defendants’ compelled statements. Govt’s June 30, 2009 Letter at 5.

Although the prosecutors redacted some tainted material from the transcripts before presenting them to the second grand jury, Defs.’ Kohl Ex. 46 at 10-12, the efficacy of these efforts was necessarily limited by the fact that the prosecutors and taint attorneys could only identify taint apparent on the face of the transcripts. The shortcomings of this process were exposed at the Kastigar hearing, during which the defendants identified several instances in which the prosecutors failed to redact — and therefore presented to the second grand jury — portions of testimony from Frost and Murphy that had been shaped by their exposure to the defendants’ compelled statements.

The starkest examples concern the testimony regarding defendant Slatten. Frost wrote in his September 18, 2007 statement that at the beginning of the shooting, he heard “two pops” coming from either the command vehicle or the follow vehicle, which he assumed were “pen flares,” a type of warning flare fired as a signal to nearby vehicles not to approach. Defs.’ Frost Ex. 7; Hr’g Tr., Oct. 21, 2009 p.m. at 11-12. When he appeared before the first grand jury on November 27, 2007, however, Frost testified that although he had originally thought that the two pops had been pen flares, “if [he] had to guess right now, [he would] say that those two pops were from the sniper rifle out of the command vehicle.” 39 GJ Ex. 89 at 60. Frost would again testify before the first grand jury that although his conclusion was not based on any damage that he saw, and although he never saw Slatten fire his weapon, the two pops “sounded like a 308 [sniper] rifle” and that “the sniper rifle is what [he thought] it was.” GJ Ex. 90 at 25. Again, Frost would later testify, “[b]ased on the sound, I think it was his — I think it was the rifle,” referring to Slat-ten’s sniper rifle. Id. at 90. None of this testimony was redacted from the transcript that was presented to the second grand jury. GJ Ex. 89 at 60; GJ Ex. 90 at 25, 90.

Likewise, in the first moments of the Nisur Square shooting, Murphy also heard several “loud pops” and immediately turned to others in his vehicle to ask if they were the sound of pen flares. 40 Hr’g Tr., Oct. 14, 2009 p.m. at 91-93. Yet Murphy testified before the first grand jury that although he could not be certain, he believed the two loud pops were shots *146 fired from Slatten’s sniper rifle. GJ Ex. 94 at 22. This testimony was also presented to the second grand jury. 41 See id

Thus, in the immediate aftermath of the incident, neither Frost nor Murphy indicated that he believed that the loud pops were shots fired by defendant Slatten and, in fact, both witnesses’ immediate reaction was that the loud pops were likely pen flares. Yet when they testified before the first grand jury two months later, both men testified that they believed the two loud pops were shots fired by Slatten from his sniper rifle.

Both Frost and Murphy acknowledged reading the sworn written statement given by Slatten to State Department investigators, in which he acknowledged firing his sniper rifle at threats he allegedly perceived at Nisur Square. Hr’g Tr., Oct. 14, 2009 p.m. at 6-7, 55; Hr’g Tr., Oct. 21, 2009 a.m. at 82; Hr’g Tr., Oct. 21, 2009 p.m. at 26. Thus, the evidence overwhelmingly suggests that the evolution of Frost and Murphy’s recollections of the events at Nisur Square resulted from their exposure to Slatten’s written statement. 42

These witnesses’ inability to disaggregate what they had personally witnessed from what they later read about the incident was further evidenced by Murphy’s testimony regarding defendant Slough. Murphy testified before the first grand jury that immediately after he heard the “two pops,” he “heard automatic weapons fire coming from [he] believe[d] Paul Slough, but [he] wasn’t looking back at the time.” GJ Ex. 94 at 23. At the Kastigar hearing, Murphy acknowledged that his testimony may have been influenced by his exposure to Slough’s written statement:

Q: [A]t the time you gave this testimony you had already read Paul Slough’s sworn statement?

A: I believe so.

Q: And you knew from that sworn statement that he acknowledged firing at the white Kia at the outset of the incident, didn’t you?

A: I would have, yeah.

Q: So this testimony where you represented you believed it was Paul Slough firing, but you didn’t see who was shooting it, was influenced in part by what you read; isn’t that fair to say?

A: Yeah, that would be fair to say.

Hr’g Tr., Oct. 14, 2009 p.m. at 115.

On redirect examination, Murphy reiterated that he assumed that defendant Slough had fired the initial burst of automatic gunfire based on the direction from which he heard the shots coming. Id at 125. But this fact underscores the scope of impermissible use that Kastigar prohibits. For even if Murphy’s first impression as to the source of the initial burst of automatic gunfire was based solely on his perception of the incident, his testimony at the Kastigar hearing indicates that he found confirmation for that initial impression in Slough’s own account of the incident, and that this influenced his testimony that he “believed” that Slough had fired *147 the initial burst of automatic gunfire at the white Kia. Id. at 115; GJ Ex. 94 at 23. This constitutes a squarely impermissible evidentiary use of Slough’s compelled statement. See North I, 910 F.2d at 861 (stating that “[i]f the government uses immunized testimony to refresh the recollection of a witness (or to sharpen his memory or focus his thought) when the witness testifies before a grand jury ... then the government clearly has used the immunized testimony”).

As it did with the tainted testimony regarding Slatten, the prosecution presented Murphy’s tainted testimony from the first grand jury to the second grand jury. See GJ Ex. 94 at 23. This occurred because the taint was not apparent on the face of the transcript and because the prosecution made no effort before presenting Frost and Murphy to the first grand jury to ascertain the extent to which they could segregate their personal recollections from the accounts they had read. 43 Accordingly, the court concludes that the government made impermissible evidentiary use of the compelled statements of Slat-ten and Slough to obtain the indictment against them.

Moreover, the evidence described above casts substantial doubt on the assertions of Murphy and Frost that they were able to segregate what they had seen and heard from what they had read of all of the defendants’ compelled accounts. Although the government emphasizes that Frost and Murphy personally witnessed many of the events about which they testified, Govt’s Mem. at 29-32; Govt’s Post-Hr’g Reply Mem. (“Govt’s Reply Mem.”) at 20, this fact would be insufficient to satisfy the government’s burden, as Kastigar bars the use of immunized statements to shape in any way the evidence presented to the grand jury. See Hylton, 294 F.3d at 134 (observing that the government “must demonstrate that witnesses who testified against the defendant were not influenced — ‘shaped, altered, or affected’ — by that information”); United States v. Poindexter, 951 F.2d 369, 376 (D.C.Cir.1992) (holding that it is the government’s burden “to demonstrate that the immunized testimony did not influence the witness”) (emphasis omitted); North II, 920 F.2d at 942 (observing that Kastigar is violated “whenever the prosecution puts on a witness whose testimony is shaped, directly or indirectly, by compelled testimony”). The evidence described above reveals the manner in which the defendants’ compelled statements could (and did) influence these witnesses’ testimony in ways not apparent on the face of the transcript. And even if the government could identify specific areas of testimony that were demonstrably free from taint, this fact would hardly establish that the witnesses’ exposure did not influence other areas of their testimony. See Poindexter, 951 F.2.d at 376 (observing that “where a substantially exposed witness does not persuasively claim that he can segregate the effects of his exposure, the prosecution does not meet its burden merely by pointing to other statements of the same witness that were not themselves shown to be [Jtainted”).

Given the fact that Frost and Murphy were exposed to dozens of media accounts based on all of the defendants’ compelled statements, the fact that the prosecutors *148 identified numerous instances in which Frost and Murphy testified based on the defendants’ statements (despite having been admonished not to do so), and the fact that the defendants identified multiple other instances in which tainted testimony was presented to the second grand jury because the taint was not readily apparent from the transcript of the first grand jury, the court finds unpersuasive the claims that Frost and Murphy were able to segregate the effects of their exposure to the defendants’ compelled statements. In addition, the court concludes that the haphazard measures taken by the government to salvage this testimony before presenting it to the second grand jury were sorely deficient, as was starkly demonstrated at the Kastigar hearing. Therefore, the government has failed to carry its burden of demonstrating by a preponderance of the evidence that the compelled statements of the defendants, to which Frost and Murphy were concededly exposed, did not in any way influence their testimony as it was presented to the second grand jury.

Although it is not necessary for the court to assign fault to find that a Kastigar violation occurred, see North II, 920 F.2d at 942, the court notes that the government could have taken certain common sense precautions before presenting its case to the first and second grand juries to mitigate the risk posed by the taint issue. The prosecutors and investigators were, after all, aware from the very moment they were assigned to this case in late September 2007 that there was a significant taint problem in this case, one that prompted the entire Criminal Division of the Justice Department to recuse itself from this prosecution. See Govt’s May 29, 2009 Letter at 2. The prosecutors and investigators could have memorialized any relevant testimony at the outset of the investigation to create a record establishing the absence of taint. Cf North I, 910 F.2d at 943 (observing that if the prosecution fails to “can” a witness’s testimony prior to his or her exposure to immunized testimony, “it may well be extremely difficult for the prosecutor to sustain its burden of proof that a witness exposed to immunized testimony has not shaped his or her testimony in light of the exposure, or ... been motivated to come forward and testify in light of the immunized testimony”). Failing that, the government could have directed taint attorneys to meet with the potentially tainted witnesses before they testified at the grand jury to probe the extent to which they could testify based on personal recollection and identify areas in which their recollections were intertwined with information gleaned from the defendants’ compelled statement. See id. at 863 (noting that “[a] central problem in this case is that many grand jury and trial witnesses were thoroughly soaked in [the defendant’s] immunized testimony, but no effort was made to determine what effect, if any, this extensive exposure had on their testimony”). At a bare minimum, the government’s trial team could have advised the witnesses in October 2007 not to seek out press reports containing the defendants’ compelled accounts to DSS investigators. The prosecutors and investigators in this case took none of these precautions and the result was a clear violation of the defendants’ constitutional rights. Absent a showing of harmlessness beyond a reasonable doubt, the dismissal of the indictment is the only permissible result.

b. The Frost Journal Was Tainted

During both grand jury proceedings, the government relied on a document prepared by Frost titled “16 September 2007: G87 traffic circle (A true account of the events as witnessed on my part)” (“the Frost Journal”). GJ Ex. 1 at 39-47; see generally Govt’s Ex. 51. The document is *149 comprised, of nine single-spaced typewritten pages in two parts: the first dated September 21, 2007, and the second dated October 3, 2007. See generally Govt’s Ex. 51. The parties did not submit evidence concerning how long it took Frost to compose each part of the document or if and when Frost revised either portion of the document. See Hr’g Tr., Oct. 21, 2009 a.m. at 88-94; Hr’g Tr. Oct. 21, 2009 p.m. at 12-14, 40-42.

The Frost Journal contains a detailed account of the Nisur Square incident, told from Frost’s perspective, including the events leading up to and immediately following the shooting. See generally Govt’s Ex. 51. The government takes the position that the Frost Journal constitutes an independent, untainted account, Govt’s Mem. at 31; Govt’s Reply at 20-21, while the defendants assert that the document was tainted because it was drafted in response to the defendants’ compelled statements, Defs.’ Mem. at 14-15.

The Circuit has made clear that Kastigar prohibits prosecutors from using immunized testimony to influence a witness’s decision to testify. See Hylton, 294 F.3d at 134 (stating that it is “rather obvious, under Kastigar and North, that if Hylton’s [immunized] statements were a cause of Wright’s decision to plead and testify against Hylton, Wright’s testimony was impermissible even if the government had prior knowledge of Wright’s role”); see also North I, 910 F.2d at 864 (noting that the testimony of Robert McFarlane, a key government witness, constituted an impermissible evidentiary use because his testimony was motivated by and directly responded to immunized testimony given by the defendant); North II, 920 F.2d at 942 (holding that “even where a witness testifies from personal knowledge, use within the meaning of Kastigar may occur ... if the immunized testimony influenced the witness’ decision to testify”).

The Frost Journal begins with four paragraphs explaining Frost’s motivation for authoring the account:

It is now the 21st of September and I feel that I must put my eyewitness account of what I have seen down on paper, so that as time goes further on my recollection is not diminished on those sad events.

Since that Sunday I have read numerous articles on the internet about the events of that day. From the N.Y. Times, LA Times, AP, UPI, Reuters, and various other outlets. I have seen on various forum discussions on what happened based on those news stories and the obligatory opinions offered with the no real idea of what truly happened out there.

Many times I have hit the respond key and typed my message out only to stop, rethink and delete what I had written. As of now, 5 days after the event, it seems that the [State Department] and [Blackwater] are locked into their stories and the real story will forever stay shrouded from the public[.]

Govt’s Ex. 51 at 1.

The foregoing indicates that at least two motivations compelled Frost to write the journal. The first was his desire to preserve his recollection. See id. The second, however, was his desire to respond to news reports reflecting the false “stories” of the State Department and Blackwater; he devotes three out of four paragraphs of the introduction to describing the latter. See id.

Frost’s testimony at the Kastigar hearing confirmed the dual motivations underlying his creation of the journal. He testified that the “main reason” he wrote the account was to memorialize the events to *150 “keep everything straight.” Hr’g Tr., Oct. 21, 2009 a.m. at 90; Hr’g Tr., Oct. 21, 2009 p.m. at 41. He also, however, acknowledged that part of his motivation resulted from his review of news articles which he felt were inaccurate and his desire to set the record straight. Hr’g Tr., Oct. 21, 2009 p.m. at 41.

A review of the media accounts published between September 16, 2007 and September 21, 2007 reveals that the State Department’s “story” was that Raven 28 took action after coming under small arms fire, 44 an account that was clearly based on the content of the defendants’ compelled statements to State Department investigators. 45 For instance, on September 17, 2007, Time/CNN reported that it “had obtained an incident report prepared by the U.S. government describing the firelight” which stated that “the motorcade was engaged with small arms fire from several locations” and that the convoy “returned fire to several identified targets” before leaving the area. Defs.’ Media Ex. 23. On September 19, 2007, ABC News produced a report stating that officials were investigating “how many, if any, bullet holes there are in the Blackwater vehicles, to confirm the Blackwater guards’ version of events that they were fired upon.” Defs.’ Media Ex. 28. Likewise, on September 21, 2007, the Los Angeles Times published a statement from a State Department spokesperson indicating that the Raven 23 convoy “came under attack, and there was defensive fire as a result of that.” Defs.’ Media Ex. 40. This evidence indicates that the Frost Journal was prepared, at least in part, as a response to the accounts given by the defendants and other Raven 23 members, which were conveyed to the media through the State Department. 46

The government points out that Frost demonstrated a desire to give his account of the incident before he was exposed to any such media accounts; in the immediate aftermath of the incident, Frost contacted Wainscott, his superior in Raven 23, to tell him that he felt uncomfortable about the things he had seen, and met with other members of the convoy who also had con *151 cerns about what had transpired. Hr’g Tr., Oct. 21, 2009 a.m. at 71-75. Indeed, the court has no doubt that Frost’s feelings about the events of September 16, 2007 instilled in him a desire to tell his story and may well have played a role in his decision to write the Frost Journal. See id. at 90-91.

Yet the Circuit has never suggested that Kastigar applies only in circumstances in which a witness’s exposure to immunized testimony was the sole or principal cause of his decision to provide inculpatory evidence against the defendant. To the contrary, the Circuit has made clear that a Kastigar violation occurs whenever exposure to immunized testimony was “a cause” of the witness’s decision to testify, see Hylton, 294 F.3d at 134, and that impermissible use within the meaning of Kastigar occurs when the immunized testimony “influences” the witness’s decision to testify, see North II, 920 F.2d at 942. This much is easily established by the opening paragraphs of the Frost Journal and his testimony at the Kastigar hearing. Because the defendants’ compelled statements were a cause of Frost’s decision to draft the Frost Journal, it is tainted and its use in the grand jury was impermissible. 47

In the October 3, 2007 portion of the Frost Journal — written after the contents of the defendants’ September 18 written statements appeared in news reports and after Slough’s written statement had been posted to the Internet — Frost expressed his anger that the defendants had not, in his view, given truthful accounts to DSS investigators:

“You guys did nothing wrong out there, and we’ve been completely honest with the State Department with their investigation[,]’ [t]he [Blackwater manager] stated. I thought to myself, you [guys] have been anything but honest with the State Department and their investigation .... We’ve been praised ... for being honest, coming forward and ‘doing the right thing’, however they haven’t done the right thing by telling the State Department what they knew.... I left the office with a rage building inside me.

Govt’s Ex. 51 at 8.

This passage makes clear that the October 3, 2007 portion of his account was motivated by his exposure to the defendants’ compelled statements. Moreover, it casts significant doubt on Frost’s cooperation with prosecutors. The meeting described in this portion of the Frost Journal took place immediately before the FBI came to conduct interviews of the Raven 23 team members, including Frost. Hr’g Tr., Oct. 21, 2009 a.m. at 94-95. The *152 government relied heavily on Frost because during his initial FBI interview, Frost left Agent Patarini with the “impression that [he] had ... a story that he would like to tell, and under the right circumstances he would tell that to us.” 48 Hr’g Tr., Oct. 21, 2009 p.m. at 104-05. The evidence thus suggests that as with McFarlane in North I, Frost’s desire to cooperate with investigators, communicated to Agent Patarini during his initial interview, was motivated by his desire to correct what he saw as falsehoods in the defendants’ compelled accounts. This too reflects impermissible use of the defendants’ compelled statements.

Moreover, the government failed to establish at the Kastigar hearing that the defendants’ compelled statements did not affect the content of the account contained in the Frost Journal. Media reports reflecting the defendants’ compelled written statements began appearing on September 28, 2007. See Defs.’ Media Ex. 2. Frost acknowledged reading the sworn written statements of Slatten and Slough, and was exposed to numerous media accounts regarding the Nisur Square shooting. Hr’g Tr., Oct. 21, 2009 a.m. at 82-83. Although it appears that Frost began writing his account before the defendants’ compelled statements appeared in the news media on September 28, 2007, there was no testimony or other evidence indicating that he completed the September 21, 2007 portion of the journal, which spans seven single-spaced pages, prior to his exposure to the defendants’ compelled statements. See id. at 64-98; Hr’g Tr., Oct. 21, 2009 a.m. at 88-94; Hr’g Tr. Oct. 21, 2009 p.m. at 12-14, 40-42; Govt’s Ex. 51 at 1-7. Nor did the government present any evidence that Frost did not alter or amend the Frost Journal at any time after his exposure. See Hr’g Tr., Oct. 21, 2009 a.m. at 64-98; Hr’g Tr., Oct. 21, 2009 p.m. at 1-61.

This does not represent some “abstract possibility of taint.” See United States v. Byrd, 765 F.2d 1524, 1529 (11th Cir.1985) (holding that the district court erred in finding a Kastigar violation based on the mere “possibility” that the investigating agent had used the defendant’s immunized testimony, despite the fact that all of the evidence relied on by the government had been obtained prior to the time the defendant gave his immunized testimony). Frost was exposed to the defendants’ compelled accounts during the time he was writing the Frost Journal. See Govt’s Ex. 51 at 1-9; Hr’g Tr., Oct. 21, 2009 a.m. at 82-83. Those compelled statements dealt with precisely the same events about which he was writing. 49 See generally Govt’s Ex. 51; Slough Stmt. (Sept. 18, 2007); Slatten Stmt. (Sept. 18, 2007); Ball Stmt. (Sept. 18, 2007); Heard Stmt. (Sept. 18, 2007); Liberty Stmt. (Sept. 18, 2007). Furthermore, it is clear from the October 3, 2007 portion of the Frost Journal that the falsity of the defendants’ statements to the DSS investigators had a powerful effect on Frost. See Govt’s Ex. 51 at 8-9. Under these circumstances, it was clearly the government’s burden to show that this account was not affected by Frost’s exposure to the defendants’ compelled state *153 ments. See Hylton, 294 F.3d at 134. The government’s failure to meet this burden leaves the court with no choice but to find yet another impermissible use of compelled testimony.

c. The Information Provided by Cooperating Witness Ridgeway Was Tainted

The government also relied heavily on information provided by cooperating witness Ridgeway to obtain the indictment against the defendants. The prosecutors provided the second grand jury with Ridgeway’s factual proffer in support of his guilty plea, which contained a detailed account of the Nisur Square shooting and characterized the incident as an unprovoked and unjustified assault on innocent civilians. See GJ Ex. 1 at 32-38; Govt’s Ex. 311 (Grand Jury Tr., Nov. 20, 2008 a.m. at 63-72). The Ridgeway factual proffer devoted particular attention to the engagement of the white Kia, concluding that “a reasonable person ... would have recognized that the white Kia did not exhibit any indicia of a VBIED attacker as provided in the Missions Firearm Policy.” GJ Ex. 1 at 37.

The government also presented the second grand jury (through Agent Powell) 50 with statements given by Ridgeway to the FBI that provided detailed allegations against all of the defendants other than Slatten. Govt’s Ex. 311 (Grand Jury Tr., Dec. 2, 2008 p.m. at 12-13). Through Agent Powell, Ridgeway informed the second grand jury that Slough had fired an M203 grenade at the white Kia and had fired at other locations around Nisur Square, id. at 13-14, that Liberty admitted firing his weapon out of his driver’s porthole, id. at 31-32, that Heard fired his machine gun at various targets to the south of the circle and also fired a grenade, id. at 34-36, and that Ball admitted firing two shots at the white Kia sedan, id. at 44.

The defendants posit that by declining to call Ridgeway to testify at the Kastigar hearing, the government failed to carry its burden and the indictment must be dismissed in its entirety. Defs.’ Mem. at 8-12. They contend that North I prohibits the court from relying on mere warnings and hearsay to conclude that the government has met its Kastigar burden, and requires the government to submit every witness who testified at the grand jury to cross-examination at a Kastigar hearing. 51 Id. at 11-12.

The government asserts that it may meet its Kastigar burden by relying on a combination of different types of evidence, including live witnesses, documents and hearsay testimony. Govt’s Reply Mem. at 5. The government notes that the court denied the defendants’ objections to the introduction of hearsay testimony to establish the absence of taint for both Ridgeway and the Iraqi witnesses, on the grounds that hearsay evidence is admissible at a pretrial suppression hearing. 52 Id. at 4; *154 Hr’g Tr., Oct. 19, 2009 p.m. at 47; Hr’g Tr., Nov. 2,2009 a.m. at 14.

Although North I demands that the Kastigar inquiry proceed “witness by witness,” NoHh I, 910 F.2d at 872-73, it is unclear whether the Circuit intended to adopt a categorical rule that in every case, the government must present every potentially exposed grand jury witness for live testimony at a Kastigar hearing. North I and North II make clear that the district court may not substitute an in camera review for an adversarial hearing, during which the defendant may probe whether the government has met its Kastigar burden. See id.; North II, 920 F.2d at 943-44. Yet the Circuit has emphasized the flexibility to be afforded the government in attempting to meet its Kastigar burden, noting that the court has never meant to suggest “that the prosecutor was barred from trying to show in any fashion that a witness’ testimony was not influenced by the immunized testimony.” North II, 920 F.2d at 943 (emphasis added) (stating also that the court has never intended to preclude the government’s use of “any techniques” to meet its burden). 53 On the other hand, North II indicates that one of the deficiencies of the in camera review relied on by the district court was that “this ‘review’ neatly avoided any cross-examination of witnesses who were admittedly exposed to the immunized testimony.” Id. at 943-44 (deeming the notion that the district court’s in camera review could take the place of an adversarial hearing “quite a remarkable proposition”).

Ultimately, the court need not resolve this issue because even assuming that the government may meet its Kastigar burden through documentary evidence and hearsay and without the live testimony of the potentially tainted witness, the evidence offered by the government with respect to cooperating witness Ridgeway failed to establish that it made no use of the defendants’ compelled statements when it presented evidence from Ridge-way to the second grand jury. As an initial matter, it is undisputed that Ridge-way was exposed to the defendants’ compelled statements. The government has acknowledged that Ridgeway saw the September 18 written statements of Slatten and Heard soon after the incident, and that he encountered Slough’s September 18 written statement in media reports in November 2007. Govt’s June 30, 2009 Letter at 9. Ridgeway also received the October 2, 2007 e-mail from Randall containing a link to an ABC News televised report which purported to based on all nineteen Raven 23 members’ accounts. Defs.’ Murphy Ex. 9. In addition, Ridge-way read press coverage and participated in Internet chat rooms that contained references to defendant Slough’s sworn statement. See Govt’s Ex. 312.

The evidence indicates that Ridgeway’s exposure to these statements was not fleeting, but may very well have affected his recollection of the events at Nisur Square. During a December 2008 interview with the prosecution team, Ridge-way’s attorney advised Kohl and Malis that there were “Kastigar risks” associated with Ridgeway’s recollection of specific targets fired on by Slough. Hr’g Tr., Nov. 2, 2009 p.m. at 101-04. Malis testified that he understood Ridgeway’s lawyer to be saying that if Ridgeway provided the government with information on those topics, there was a risk he would be providing *155 information that resulted from his exposure to compelled testimony. Id. at 103.

In the face of this potential taint, the government relies solely on the following to meet its burden: its admonition to Ridgeway to limit his statements to his personal recollection, see Hr’g Tr., Nov. 2, 2009 a.m. at 13-16, 20-21; Ridgeway’s assertion that he could do so (for all the defendants other than Slatten), 54 id. at 15-16; and the fact that that Ridgeway directly observed many of the events about which he testified, Govt’s Reply Mem. at 9. Yet this Circuit has made abundantly clear that mere admonitions and conclusory assertions are insufficient to establish the absence of taint. North I, 910 F.2d at 868 (holding that “the District Court’s reliance on warnings to witnesses (to avoid testifying as to anything they had learned from North’s immunized testimony) was not sufficient to ensure that North’s testimony was not used”); see also Poindexter, 951 F.2d at 375-76 (holding that the district court’s instruction to a witness not to “go into other people’s testimony” was not sufficient to ensure the absence of taint). In addition, as previously discussed, the government’s burden is not limited to merely showing that the witness possessed an independent basis from which he might have testified. See Hylton, 294 F.3d at 134. Rather, the government must prove that the witness’s exposure to compelled statements did not — in any way — influence the testimony he gave to the grand jury. See id.\ North II, 920 F.2d at 942.

Thus, the fact that Ridgeway was present at Nisur Square is insufficient to demonstrate the absence of taint. Instead, the government was required to prove that every one of the detailed allegations that Ridgeway made against the defendants was not influenced in any way by his prior exposure to their compelled statements. See Hylton, 294 F.3d at 134; Poindexter, 951 F.2d at 376. The government has utterly failed to meet this burden, a striking failure given the significance of this witness to the overall prosecution of this case. Ridgeway acknowledged having read the sworn written statements of Slough, Slatten and Heard; he appears to have been exposed to media accounts based on the compelled statements of all the defendants; and he provided specific inculpatory testimony against every defendant except Slatten. The government’s failure to prove that it made no evidentiary use of the compelled statements through Ridgeway constitutes a clear Kastigar violation.

d. The Statements of Iraqi Witnesses Presented to the Second Grand Jury Were Tainted

The government also presented the statements of twenty-two Iraqi witnesses to the second grand jury. See generally Govt’s Ex. 1. The statements of these Iraqi witnesses were used to support the government’s allegations against every defendant. See generally GJ Ex. 107. As it did with cooperating witness Ridgeway, the government attempted to meet its *156 Kastigar burden with respect to these witnesses solely through hearsay testimony. See Hr’g Tr., Oct. 19, 2009 p.m. at 48-98; Hr’g Tr., Oct. 20, 2009 p.m. at 1-109. The defendants maintain that the government’s failure to present these witnesses for cross-examination at the Kastigar hearing necessarily requires the dismissal of the indictment, Defs.’ Mem. at 19, while the government avers that it was neither feasible nor necessary to have all the Iraqi witnesses testify at the Kastigar hearing, Govt’s Reply Mem. at 4-11.

Once again, however, the court need not reach the issue because even assuming the government could meet its heavy Kastigar burden through hearsay testimony alone, the evidence offered by the government failed to demonstrate that the defendants’ compelled statements did not in any way influence the statements given by these Iraqi witnesses. As previously discussed, the defendants’ September 18 written statements were widely reported in the media in the weeks following the incident, and reports conveying the substance of the defendants’ accounts to DSS agents — that they were responding to hostile fire — began to appear almost immediately after the incident. The government does not dispute that the Nisur Square incident was notorious in Iraq or that media coverage of the incident saturated Baghdad. See Defs.’ Mem. at 17; see generally Govt’s Mem.; Govt’s Reply. FBI Special Agent Carolyn Murphy, one of the agents who interviewed Iraqi witnesses in October 2007, acknowledged the pervasiveness of the media coverage in Iraq about the incident, stating “that there had been reports on the news about it, that people were talking about it.” Hr’g Tr., Oct. 19, 2009 p.m. at 69.

At least one Iraqi witness, Hassan Jabir Salman, made statements reflecting his exposure to the defendants’ compelled accounts. Salman, who was shot at Nisur Square, spoke to reporters from his hospital bed and stated that “[i]t is not true when they say that they were attacked. We did not hear any gunshots before they started shooting.” Defs.’ Media Ex. 43 at 1. This statement, which was widely reported, see, e.g., Defs.’ Media Exs. 39, 40, 43, appears to have been a direct response to the defendants’ compelled accounts that they had encountered incoming fire on September 16, 2007.

The FBI’s records indicate that five of the Iraqi witnesses interviewed in June 2009 acknowledged being exposed to the statements of Raven 23 members. 55 See Govt’s Ex. 23. The statements of three of these five witnesses were presented to the second grand jury. See GJ Ex. 1; see generally GJ Ex. 107. At least one such witness, Sarhan Deab Abdul Moniem Da-Zubaidi, acknowledged having been exposed to specific information from the defendants’ compelled statements through a prior interview with DSS agents. See Hr’g Tr., Oct. 19, 2009 p.m. at 94-95. In June 2009, Da-Zubaidi informed FBI investigators that when he was interviewed by DSS agents prior to his October 2007 interview with the FBI, “he had been told by investigators that the Blackwater guards had said that he was pushing the [white Kia] towards their convoy.” 56 Id. *157 at 95. Allegations made by Da-Zubaidi were specifically included in the summary of evidence against Slough and Ball. See GJ Ex. 107 at 1, 7.

At the Kastigar hearing, the government attempted to establish that all of the Iraqi witnesses on whom it relied were free from taint solely through the hearsay testimony of Agent Murphy. See Hr’g Tr., Oct. 19, 2009 p.m. at 48-98; Hr’g Tr., Oct. 20, 2009 p.m. at 1-109. In October 2007, Agent Murphy was part of a team of FBI investigators that traveled to Iraqi to conduct interviews of Iraqi eyewitnesses identified by the INP. Hr’g Tr., Oct. 19, 2009 p.m. at 64-66. Although Agent Murphy testified that she and her fellow investigators admonished the witnesses not to relay information obtained through press accounts, see id. at 69-70, notable is the fact that not a single FBI investigative report for those interviews reflects that such an admonishment was given, see Hr’g Tr., Oct. 20, 2009 p.m. at 59-62. At any rate, as previously discussed, mere admonishments are patently insufficient to demonstrate the absence of taint. See North I, 910 F.2d at 868; Poindexter, 951 F.2d at 375-76. Agent Murphy identified no other steps taken by the FBI agents in October 2007 to ensure that they were not receiving tainted information from the witnesses they interviewed. See Hr’g Tr., Oct. 19, 2009 p.m. at 66-71.

In June 2009, Agent Murphy returned to Iraq with an FBI investigative team to conduct additional pretrial interviews of certain Iraqi witnesses. Id. at 72. On June 10, 2009, Agent Murphy circulated an e-mail to the prosecution team and fellow investigators, regarding “[wjitness admonishments,” Govt’s Ex. 45, which established a three-step taint protocol for these interviews: (1) the witness was admonished to provide information only from his or her personal recollection; (2) the witness was asked whether he or she had been exposed to media accounts or the defendants’ statements; and (3) if affirmative, the witness was asked whether this exposure had influenced his or her memory or affected his or her decision to cooperate with investigators, see id. There is no evidence that the FBI interviewers conducted any type of probing inquiry into whether exposure to the defendants’ compelled statements influenced the witnesses’ recollections. Rather, it appears that if a witness answered that his or her memory was not influenced by exposure to compelled testimony, the inquiry ended. See id.; Hr’g Tr., Oct. 19, 2009 p.m. at 93-94. This perfunctory protocol falls far short of establishing that the witnesses’ testimony was not influenced in any way by their exposure to the defendants’ compelled statements.

In sum, the government failed to establish that the Iraqi witnesses it presented to the second grand jury were not in any way influenced by their previous exposure to the defendants’ compelled statements. This evidentiary use of tainted information constitutes yet another Kastigar violation. 57

*158 2. Nonevidentiary Use of the Defendants’ Compelled Garrity Statements

a. The Defendants’ Garrity Statements Prompted the Decision to Charge Defendants Heard and Ball

The defendants contend that the trial team’s exposure to the compelled statements of Heard and Ball played a central role in the government’s decision to include them as targets of the prosecution. Defs.’ Mem. at 35-36. They note that although the prosecution had identified both Heard and Ball as shooters by the end of 2007, the trial team did not add them as targets of the prosecution until after they had obtained and reviewed the defendants’ compelled statements. Id. The government insists that the decisions to charge Heard and Ball were based on the discovery of evidence in early 2008 unrelated to the defendants’ compelled statements. Govt’s Mem. at 40-41.

An examination of the evidence reveals that although the government’s trial team had identified defendant Heard as a shooter by December 2007, he was not listed as a target of the prosecution in the trial team’s internal investigative updates until March 2008, Govt’s Ex. 70, after the prosecutors had completed their interviews of the DSS agents in January and February 2008. Moreover, Kohl’s draft prosecution memorandum makes clear that Heard’s September 16 interview statement played a central role in the decision to add him as a target of the prosecution. See Defs.’ Kohl Exs. 68-71, 73. The prosecution memorandum quotes extensively from Heard’s September 16 interview statement, including his descriptions of firing machine guns at the white Kia and at an alleged insurgent. Id. Moreover, the memorandum specifically notes that after completing his interview, Heard returned to the interview room approximately forty-five minutes later of his own accord “extremely upset” and disclosed to DSS agents additional details regarding his actions, including the fact that he had fired an M203 grenade. Id. The prosecution memorandum states that Heard’s “initial reluctance” to provide full disclosure demonstrated his “consciousness of guilt” and showed that he “knew his actions [in Nisur Square] were excessive and reckless.” 58 Id.

The weight that Kohl placed on Heard’s September 16 statement is further revealed in Kohl’s specific reference to that statement when attempting to persuade Heard to cooperate with the government in the summer of 2008. Hr’g Tr., Oct. 29, 2009 a.m. at 161-64. Hulser, the head of the taint team and a senior prosecutor, testified that he did not authorize Kohl to use Heard’s September 16 statements in this manner. Hr’g Tr., Oct. 23, 2009 a.m. at 77.

Kohl testified that his decision to add Heard as a target resulted from two developments: first, his “discovery” in February 2008 that the white Kia was traveling at a slow rate of speed and thus did not pose a threat to the convoy, and second, *159 his conversation with Colonel Boslego regarding the inherent impropriety of firing an M203 grenade in an urban area. Hr’g Tr., Oct. 28, 2009 a.m. at 41-48. Notably, Frost, Mealy and Murphy had all already testified in late 2007 that the white Kia was traveling slowly, 59 and Murphy even testified that he believed it was not a threat. GJ Ex. 94 at 22. In fact, during his examination of Mealy at the first grand jury, Kohl challenged Mealy’s assertion that the engagement of the white Kia may have been justified, eliciting testimony from Mealy about the escalation of force rules and the speed at which the Kia was moving. GJ Ex. 92 at 50-54.

Pontieello testified that Heard became a target once the prosecutors decided to “charge the white Kia” and mentioned nothing about Kohl’s conversation with Colonel Boslego. Hr’g Tr., Oct. 27, 2009 p.m. at 41. Indeed, when asked to confirm that Heard’s statement to the DSS investigators caused the government to target him, Pontieello responded, “I can’t say that that’s the only basis.” Id. at 53. Thus, Pontieello acknowledged what is readily apparent from the timing of Heard’s addition to the target list, as well as from Kohl’s references to that statement in the prosecution memorandum — that Heard’s September 16 interview statement played a substantial role in the government’s decision to focus on him as a target of the prosecution.

Though there is less direct evidence regarding the decision to charge defendant Ball, the evidence indicates that statements he gave to DSS investigators played a central role in that decision as well. Ball was not listed as a target of the investigation until the investigative update of April 24, 2008. See Govt’s Ex. 70. By that time, the trial team had interviewed DSS agents regarding the compelled statements of Ball and the other defendants. In addition, it appears that in April 2008, prosecutors and investigators reviewed an unsigned draft of Ball’s September 18 written statement, which had been acquired through the execution of the search warrant based on the defendants’ September 16 oral statements. 60 Pontieello could identify no evidence against Ball obtained between the March 10, 2008 investigative update, which does not list Ball as a target of the investigation, and the April 24, 2008 investigative update, which names Ball as a target for the first time. Hr’g Tr., Oct. 27, 2009 p.m. at 31-34; see Govt’s Ex. 70.

The sole reason offered by the government to explain its decision to add Ball as a target in late April 2008 is, again, the trial team’s “discovery” in February 2008 that the white Kia was traveling at a slow rate of speed. Govt’s Mem. at 40. Yet, as-previously discussed, this asserted justification is simply not credible. Thus, the government has faded to carry its burden to show that it did not use Ball’s compelled statements to focus the prosecution on him. To the contrary, the common sense inference to be drawn from this evidence, given the timing of the decisions and the *160 government’s failure to offer a credible explanation to the contrary, is that Ball’s disclosures to DSS agents played a determinative role in the government’s decision to prosecute him. This significant and tangible nonevidentiary use of the compelled testimony of Heard and Ball is barred by Kastigar and Garrity.

b. The Defendants’ Garrity Statements Guided the Government’s Investigation and Prosecution

The defendants argue that in light of the government’s persistent efforts to obtain the defendants’ compelled statements and the value of these statements to the prosecution, it is simply implausible that the government did not make significant use of these statements to steer its investigation and to shape its view of the case. Defs.’ Mem. at 38. The government maintains that regardless of whatever exposure they may have had to the defendants’ compelled statements, the prosecutors and investigators made no investigative or strategic use of these compelled statements. Govt’s Reply Mem. at 12-17. The government asserts in its own defense that the trial team’s exposure to these statements resulted from a mere “miscommunication” between Kohl and Hulser, Govt’s Mem. at 13, and that this exposure did not reveal any information that was not already known to the government or that was of any use to the prosecution, id. at 41-45.

The evidence presented during the Kastigar hearing leaves no doubt that the trial team went to great lengths to obtain information gleaned from the defendants’ compelled statements. In January and February 2008, the trial team traveled to Iraq and interviewed all of the DSS agents involved in the Nisur Square investigation. On January 11, 2008, the trial team interviewed Agent Carpenter, who disclosed specific information concerning the content of the defendants’ compelled statements. Hr’g Tr., Oct. 19, 2009 a.m. at 30. This information included disclosures made by the defendants during their September 16, 2007 interviews. Id. This information also included descriptions of specific targets that Slough fired at north of the traffic circle, which Slough did not disclose during his September 16, 2007 interview but revealed in subsequent interviews with DSS agents. Id. at 30-35.

On January 25, 2008, the trial team interviewed Agent Lopez. Defs.’ Lopez Ex. 6. During this interview, the trial team obtained information regarding the specific threats identified by Slough in his compelled statements. Hr’g Tr., Oct. 15, 2009 p.m. at 60-61. Agent Lopez also disclosed that Heard had initially stated that he did not fire an M203 grenade, but returned later, “very upset about the incident,” and disclosed the full extent of his involvement. Defs.’ Lopez Ex. 6. In addition, Kohl obtained from Agent Lopez a copy of her interview notes from the September 16, 2007 interviews she had conducted. Govt’s June 30, 2009 Letter at 4.

The trial team asked Agent Farrington to provide them with his opinion about what had happened at Nisur Square, a view that he acknowledged was shaped by his review of the defendants’ compelled testimony during the DSS investigation, including a summary of their September 18 written statements and their post-September 18 re-interviews. Hr’g Tr., Oct. 16, 2009 p.m. at 84-85, 91-92; Defs.’ Farrington Ex. 4.

The trial team questioned Agent Motley about the re-interviews of the defendants conducted on September 20, 2007, during which the defendants pointed out on a map the various directions from which they had purportedly taken fire. Hr’g Tr., Oct. 16, 2009 a.m. at 71-72. Ponticello testified that the trial team did not seek a copy of the map marked by the defendants be *161 cause they recognized that it contained tainted information. Hr’g Tr., Oct. 27, 2009 p.m. at 63. Yet, incredibly, during the trial team’s interview of Agent Motley, the trial team provided Agent Motley with an aerial map of Nisur Square and had him circle the locations that had been identified by the defendants on the admittedly tainted map. 61 Id. at 66.

In February 2008, the trial team used information from the defendants’ September 16 interview statements to obtain search warrants for unsigned copies of the defendants’ sworn written statements. 62 Hr’g Tr., Oct. 22, 2009 p.m. at 17-21. Agent Patarini acknowledged reading the written statements of Ball and Heard and providing copies of these statements to the prosecutors. Id. at 22. The government’s only explanation for the search warrant project was that it was designed to investigate potential false statement prosecutions. See Govt’s Mem. at 15; Govt’s Ex. 288 at 9-10; Hr’g Tr., Oct. 28, 2009 p.m. at 44. But Kohl himself acknowledged that it was inappropriate for him to view the September 18 statements in connection with a false statements prosecution while still considering a prosecution based on the activities disclosed in those statements, Hr’g Tr., Oct. 28, 2009 p.m. at 70, and that the search warrant project was “close to the line, absolutely,” Hr’g Tr., Oct. 29, 2009 a.m. at 32-33.

In August 2008, months after Kohl and Hulser met to resolve the purported “miscommunication” between them regarding exposure to tainted evidence, Kohl obtained and read two “Memorandum Reports of Interviews” memorializing the statements given by the defendants during their September 16, 2007 interviews. Id. at 165-66. Kohl testified that this action was in keeping with the understanding he reached with Hulser at the April 2008 meeting. Id. at 79. But this assertion was flatly contradicted by Hulser, who testified that his position regarding the September 16 interview statements never changed and that he never authorized the trial team to obtain notes from those interviews. Hr’g Tr., Oct. 23, 2009 a.m. at 42.

It is clear from this evidence that beginning in the first months of the investigation and throughout its duration, the trial team aggressively sought out and obtained the substance not only of the defendants’ September 16 interview statements, but also of the defendants’ subsequent statements to the DSS investigators, which the government has never disputed were compelled and tainted. The trial team members immersed themselves in the defendants’ compelled statements long before obtaining the indictment in December 2008.

All of the trial team’s efforts to obtain the contents of the defendants’ compelled statements were taken in direct contraven *162 tion of the clear directives given by taint attorney Hulser. Between November 2007 and April 2008 (when Kohl and Hulser agreed that Kohl would decide what information would be sent to Hulser for screening), Hulser repeatedly advised the trial team to refrain from any contact with the DSS investigators and not to seek out information derived from the tainted DSS investigation, including the defendants’ September 16 oral statements. See, e.g., Govt’s Ex. 57; Defs.’ Kohl Exs. 9-10. Hulser never authorized the trial team’s interviews of the DSS agents in early 2008, did not approve the trial team’s efforts to obtain records and notes from the defendants’ September 16, 2007 interviews and never sanctioned the use of the defendants’ September 16 interview statements to obtain the search warrant for the defendants’ September 18 written statements. Hr’g Tr., Oct. 23, 2009 a.m. at 28-29.

The government attempts to characterize Kohl’s failure to heed Hulser’s directives as a mere “miscommunication.” Yet to accept this characterization, the court would have to accept the following: that Kohl, a seasoned and accomplished prosecutor, failed to read multiple e-mails from senior prosecutors regarding a high profile case, including one e-mail to which Kohl responded “Got it,” see Defs.’ Kohl Ex. 10; Govt’s Ex. 57; that Mullaney, a Section Chief in the DOJ National Security Division, failed to forward to Kohl numerous e-mails from Hulser containing critical information regarding the investigation, despite Mullaney’s testimony that it was his practice to forward any such emails, Hr’g Tr., Oct. 26, 2009 p.m. at 8-9; and that after his April 2008 meeting with Hulser, Kohl understood that he could use the defendants’ September 16 statements so long as the trial team’s exposure to these statements was “delay[ed] as long as possible,” Hr’g Tr., Oct. 29, 2009 a.m. at 79, an understanding diametrically at odds with Hulser’s sworn recollection of the meeting, Hr’g Tr., Oct. 23, 2009 a.m. at 42, and contrary to all of Hulser’s prior advice on the issue, see, e.g., Defs.’ Kohl Ex. 10. These inconsistent, extraordinary explanations smack of post hoc rationalization and are simply implausible. The only conclusion the court can draw from this evidence is that Kohl and the rest of the trial team purposefully flouted the advice of the taint team when obtaining the substance of the defendants’ compelled statements, and in so doing, knowingly endangered the viability of the prosecution. 63

The government argues that the trial team could not, as a matter of law, have used the defendants’ compelled statements because the statements only contained exculpatory information. Govt’s Mem. at 43-44. Yet, as the court has already observed, in each of the authorities on which the government relies to support this assertion, the determination that there was no Kastigar violation resulted from the fact that the statement was not useful to the prosecution under the circumstances of that case. See, e.g., Bartel, 19 F.3d at 1113; Caporale, 806 F.2d at 1518. In this case, by contrast, it is abundantly clear that the defendants’ compelled statements did have a value to the prosecution. De *163 fendants Slough, Slatten, Heard and Ball acknowledged having fired their weapons and specified the types of weapons they had used, the number of shots they had fired and the directions in which they had fired. See supra Part II.A.2. All of the defendants specified in great detail the locations of the various insurgent threats they had allegedly seen and specified the directions from which they had purportedly taken fire. See id. Agent Patarini testified that he may have used the information he obtained from the defendants’ statements to DSS investigators in questioning witnesses. Hr’g Tr., Oct. 22, 2009 p.m. at 70. Ponticello testified that he sought the defendants’ September 16 interview statements because he wanted to use them in the investigation and that he did make “ancillary” use of them. Hr’g Tr., Oct. 27, 2009 p.m. at 22, 77-78. Hulser, an experienced senior prosecutor, testified that although a defendant’s mere denial of involvement may be of no real use, the defendants’ statements in this case specified who they were shooting at and who was shooting back at them, and all of this was “valuable information.” Hr’g Tr., Oct. 23, 2009 a.m. at 30-32. As Hulser testified, “[i]f I were conducting an investigation without any taint at all, I would expect to have those statements, I would want to have those statements.” Id. at 32.

In the face of the unrefuted evidence that the trial team risked the entire prosecution in aggressively seeking out the defendants’ compelled statements, which provided a wealth of valuable information, the government asks this court to credit the conclusory assertions of Kohl and the rest of the trial team that they made no use of these statements to further the prosecution. Govt’s Mem. at 14, 41. The government asks too much. See United States v. Sehmidgall, 25 F.3d 1523, 1528 (11th Cir. 1994) (observing that “a government agent’s bare denials that he used the immunized testimony are insufficient to meet the government’s burden”); United States v. Harris, 973 F.2d 333, 336 (4th Cir.1992) (noting that “the government’s mere representations” that it made no use of immunized testimony “standing alone are generally insufficient to carry its burden”); Hsia, 131 F.Supp.2d at 209 (observing that “courts have universally held ... the government’s protestation that the immunized testimony did not affect its prosecution of the immunized witness to be insufficient, no matter how sincere”). It simply defies common sense that the prosecution would go to such incredible lengths to obtain the defendants’ compelled statements, flouting the advice of the taint team and taking actions that even Kohl acknowledged came “close to the line,” and then make no use whatsoever of the fruits of their efforts. 64

These facts do not describe a case of fleeting exposure late in the game that may have tangentially affected the trial team’s thought processes. See North I, 910 F.2d at 859. Rather, they reveal that the trial team went to great lengths and knowingly took great risks, at the early stages of the prosecution, to obtain statements that provided a wealth of information valuable to the prosecution. Given the prosecution’s early, ongoing and intentional immersion in the defendants’ compelled statements, the government bore the burden of demonstrating that it made no significant nonevidentiary use of the *164 defendants’ statements. See Harris, 973 F.2d at 336 (holding that the prosecution failed to meet its Kastigar burden because its failure to follow reliable taint procedures left it “unable to eliminate the significant possibility that it used ... compelled testimony as an ‘investigatory lead’ ”); cf. North I, 910 F.2d at 860 (concluding that the prosecution “could not have made significant nonevidentiary use” of the tainted information because there had been no significant exposure to that information); Barker, 542 F.2d at 484 n. 9 (suggesting that holding of McDaniel — that any nonevidentiary use was impermissible — applied only in the limited circumstance in which immunized testimony “had been obtained at a very early stage in the investigatory process”). The government’s utter failure to meet this burden requires dismissal of the indictment against all the defendants. See North I, 910 F.2d at 859-60; Hsia, 131 F.Supp.2d at 201-02 (dismissing the indictment based on the prosecutions’ failure to demonstrate that it made no nonevidentiary use of the defendant’s compelled statements).

c. The Defendants’ Garrity Statements Tainted the Physical Evidence Recovered from Nisur Square by DSS Investigators

As previously discussed, in the days following the Nisur Square incident, the DSS agents conducted a search of the traffic circle and collected physical evidence. See supra Part II.A.2. The defendants contend that the prosecution’s use of physical evidence derived from the DSS investigation is tainted. Defs.’ Mem. at 32-34. The government responds that the DSS search for physical evidence was not guided by the defendants’ compelled statements. Govt’s Reply Mem. at 16. Furthermore, the government contends that in light of the multiple, extensive searches conducted by the U.S. military and the FBI, any evidence recovered by the DSS agents would inevitably have been discovered independent of the DSS search. Id.

The evidence strongly indicates that the physical search conducted by DSS investigators was guided by the defendants’ compelled statements. On September 20, 2007, DSS investigators interviewed the four defendants who had acknowledged firing their weapons to focus on “the specific details” of the written accounts they had given on September 18, 2007. See Defs.’ Carpenter Ex. 9. Agent Farrington testified that Agent Carpenter directed them to obtain these “specific details” to assist the DSS agents in the physical search of the Nisur Square traffic circle conducted later that day. 65 Hr’g Tr., Oct. 16, 2009 p.m. at 22-23. Agent Motley informed the trial team that the physical search of the Nisur Square area was based on the defendants’ oral statements to investigators on September 20, 2007, and was designed to test the accounts given by the defendants. Hr’g Tr., Oct. 16, 2009 a.m. at 73. And although Agent Carpenter testified that there were no areas of Nisur Square that his agents would not have searched but for the information provided in the September 16, 2007 interviews, Hr’g Tr., Oct. 19, 2009 p.m. at 38, he clearly stated in his investigative report that “[f]rom [Raven] 23 statements we obtained information concerning their direction of fire and opposing fire. We searched in these areas and col *165 lected what appeared to be AK-47 shell casings and material from a white 4 door sedan. Photographs of the area were taken to include bullet strikes and the AK-47 shell casings that were collected.” Id. Accordingly, the court discerns that the physical evidence derived from the DSS investigation of the Nisur Square environment was tainted by the defendant’s compelled Garrity statements.

As for the government’s claim of inevitable discovery, the court notes that Nisur Square was not sealed as a closed crime scene between the shooting on September 16, 2007 and the arrival of FBI investigators several weeks later. Defs.’ Mem. at 33. Agent Scollan testified that when he visited the scene shortly after the incident, he noticed an Iraqi man, possibly a police officer, who appeared to be walking through the scene and picking up objects that may have been evidence. Hr’g Tr., Oct. 15, 2009 a.m. at 99-100. Agent Scollan testified that the scene “was extremely clean — picked clean as if it was being groomed for a garden.” Id. at 99. Under these circumstances, the court deems the government’s claims of inevitable discovery unpersuasive.

The government does not dispute that the physical evidence recovered from the scene was passed on to the prosecutors and investigators. See Defs.’ Mem. at 33; Govt’s Reply Mem. at 15-16; Govt’s Ex. 275. Moreover, the DSS search was aimed at testing the veracity of the defendants’ accounts and was guided by the specific details disclosed in those accounts, see Hr’g Tr., Oct. 16, 2009 p.m. at 22-24, which would suggest that the evidence recovered through this search may have been highly relevant to the criminal case eventually brought against the defendants. Indeed, the government has never argued otherwise. The trial team’s exposure to this potentially significant physical evidence triggers the government’s obligation to demonstrate that it has made no significant use of this evidence. See North I, 910 F.2d at 860; Hsia, 131 F.Supp.2d at 201-02. The government’s failure to do so constitutes yet another Kastigar violation.

3. The Government’s Myriad Kastigar Violations Cannot Be Excused as

Harmless Error

Before dismissing an indictment based on Kastigar violations, the court must determine whether the government’s use of immunized testimony was harmless beyond a reasonable doubt. See Ponds, 454 F.3d at 328-29 (citing North I, 910 F.2d at 854). “Where ... immunized evidence emerges early in the investigation, the court must determine whether the government ‘would have taken the same steps entirely apart from the motivating effect of the immunized testimony.’ ” Id. (quoting United States v. Nanni, 59 F.3d 1425, 1433 (2d Cir.1995)). “The government cannot escape its error simply by showing the availability of ‘wholly independent’ evidence from which it might have procured indictment or convictions had it not used the immunized testimony,” but must instead demonstrate beyond a reasonable doubt that the prosecution would have been vigorously pursued, and the same investigative steps taken, had the government not relied on immunized material. Id. (quoting United States v. Pelletier, 898 F.2d 297, 303 (2d Cir.1990)). “Unless the government’s use of Kastigar evidence, in light of evidence obtained from independent sources, was so unimportant and insignificant and has so little, if any, likelihood of having changed the result of the proceeding that it may be deemed harmless,” the violation of the defendant’s constitutional rights cannot be excused as harmless beyond a reasonable doubt. Id. (citing United States v. Gallo, 859 F.2d 1078, 1082 (2d *166 Cir.1988)) (internal quotation marks omitted).

Just a few examples suffice to show that the government has utterly failed to make this demanding showing as to any defendant. Defendant Slough’s compelled statements influenced the grand jury testimony of Murphy and likely tainted the information provided by cooperating witness Ridgeway and many Iraqi witnesses to the second grand jury. Defendant Slat-ten’s compelled statements shaped the grand jury testimony of Frost and Murphy, whose testimony comprised the near entirety of the evidence presented against him at the second grand jury. See GJ Ex. 107 at 6-7. The compelled statements of defendants Ball and Heard played a determinative role in the government’s decision to focus the prosecution on them. The compelled statement of defendant Liberty and the other defendants motivated the creation of the Frost Journal, influenced Frost’s decision to cooperate with the government, guided the questioning of witnesses during the investigation, resulted in the discovery of physical evidence and provided the government with a version of events that each defendant was locked into and helped steer the investigation. 66 In sum, far from being unimportant and insubstantial, the defendants’ compelled statements pervaded nearly every aspect of the government’s investigation and prosecution, and the government’s use of those statements appears to have played a critical role in the indictment against each of the defendants. Accordingly, the court declines to excuse the government’s reckless violation of the defendants’ constitutional rights as harmless error.

IV. CONCLUSION

Before the beginning of jury deliberations, a judge instructs the jury that it must perform its duty to deliberate “without prejudice, fear, sympathy or favoritism.” A judge has a concomitant obligation. When a judge, upon close examination of the procedures that bring a criminal matter before the court, concludes that the process aimed at bringing the accused to trial has compromised the constitutional rights of the accused, it behooves the court to grant relief in the fashion prescribed by law. Such is the case here.

For the reasons discussed in this Memorandum Opinion, the court grants the defendants’ motion to dismiss the indictment. 67 An Order consistent with this Memorandum Opinion is separately and contemporaneously issued this 31st day of December, 2009.