Johnson v. United States

Notice: This opinion is subject to formal revision before publication in the Atlantic and Maryland Reporters. Users are requested to notify the Clerk of the Court of any formal errors so that corrections may be made before the bound volumes go to press. DISTRICT OF COLUMBIA COURT OF APPEALS No. 24-CF-0525 CHANCE CATO JOHNSON, APPELLANT, V. UNITED STATES, APPELLEE. Appeal from the Superior Court of the District of Columbia (2023-CF2-004608) (Erik P. Christian, Judge) (Submitted January 13, 2026 Decided September 3, 2026) Mindy Daniels was on the brief for appellant. Edward R. Martin, Jr., United States Attorney at the time the brief was filed, and Chrisellen R. Kolb, Mark Hobel, Ariel Lieberman, Nicholas Reck, and Bryan H. Han, Assistant United States Attorneys, were on the brief for appellee. Before EASTERLY and SHANKER, Associate Judges, and RUIZ, Senior Judge. EASTERLY, Associate Judge: Law enforcement officers have a first-line responsibility to assess the reliability of information from confidential sources before acting on a tip. But the ultimate determination whether that information is sufficiently reliable to justify a search or seizure under the Fourth Amendment is for  2 the court to make. Police officers’ conclusory assertions and unsupported assumptions are unhelpful to that assessment. Rather, the court must be apprised of sufficient facts to allow it to independently determine whether the police complied with the Fourth Amendment. In this case, the police stopped Chance Cato Johnson on the street solely based on information from a confidential informant about a man with a gun. But the officer who worked with the informant and received the tip did not testify at the suppression hearing. The officer who did testify could only relate that the non-testifying officer had said the confidential informant was reliable and was unable to provide the court with any facts from which the court could evaluate that characterization, much less make its own reliability determination. The tip itself included only an undetailed assertion that the informant had “seen” a man with a gun and a generic description of the man’s clothing, with no information about height, weight, race, age, or hairstyle. And the government’s evidence that this description was promptly and specifically corroborated was itself weak, resting on (1) the discovery of Mr. Johnson some unknown time after the receipt of the tip at a location a couple of blocks away from where the informant said the man with the gun would be found and (2) a screenshot of multiple individuals, at least two of whom might have been the man described, that the government never meaningfully tied to the informant’s tip. Taken together, the information presented by the government regarding the tip  3 was not sufficient for the trial court to determine that Mr. Johnson’s seizure by the police was based on reasonable articulable suspicion. Because Mr. Johnson’s firearm-related convictions were based on the fruits of his illegal seizure, we reverse. 1 I. Facts and Procedural History At the suppression hearing, the government presented only one witness, Metropolitan Police Department (MPD) Officer Conner Johnson. Officer Johnson testified he was on patrol when, at “about 4:20 in the afternoon,” he “got a call from Officer Keith Cyphers regarding some information obtained by a confidential informant.” Officer Johnson testified that Officer Cyphers told him that this informant was a “reliable source that [Officer Cyphers] ha[d] been using for a while,” but Officer Johnson could not speculate whether “a while” meant “a year” or “a couple of minutes.” Even though Officer Johnson “personally d[id] not know 1 Because we agree with Mr. Johnson that the fruits of his illegal seizure should have been suppressed on Fourth Amendment grounds, we need not separately address his argument that his incriminating statements subsequent to his seizure should have been suppressed under Miranda v. Arizona, 384 U.S. 436 (1966). See Robinson v. United States, 76 A.3d 329, 342 (D.C. 2013) (explaining that, in the context of a Fourth Amendment violation, the exclusionary “rule applies to both physical evidence and testimonial evidence”). Here, the government concedes that Mr. Johnson’s admission to the police that he possessed a gun was made in response to questions that were “part of the investigatory stop.”  4 the [informant] at all,” and had not “worked with [Officer Cyphers] personally before,” Officer Johnson “took [Officer Cyphers] at his word,” without asking any other questions such as how long Officer Cyphers “had . . . been using” the informant or “why [Officer Cyphers] thought [the informant] was reliable.” Officer Johnson testified that Officer Cyphers told him that the informant had said “a man . . . with a white shirt, black shorts, blue shoes and a black bag had a firearm on his person.” Officer Johnson “th[ought]” Officer Cyphers had said this man was in the 3800 block of Minnesota Avenue. After their phone call, Officer Cyphers texted Officer Johnson a screenshot of street camera video footage, reproduced below, that Officer Johnson “believe[d]”—why, he did not say—had been taken from a private business at 3847 Minnesota Avenue and “point[ed] toward 3845.”  5 Officer Johnson “believe[d]” Officer Cyphers had viewed the video “live.” Officer Johnson could not recall whether Officer Cyphers sent any “explanatory text” with the screenshot, but he also testified that Officer Cyphers had “told” him that it had been taken “very recent[ly].” When asked “what steps [he took] to verify that the picture that Officer Cyphers [sent] . . . was relevant to the investigation [he was] undergoing,” he explained that, because Officer Cyphers had “sent it to me after we spoke,” he assumed that to be the case. Ignoring the other individuals in the screenshot, Officer Johnson described it as showing an “individual with [a] cross body bag . . . . [with] black shorts.” Officer Johnson identified that man as Mr. Johnson.  6 In response to the tip relayed by Officer Cyphers, Officer Johnson “canvassed the area.” After what he testified was ten to fifteen minutes but his body-worn camera footage documents as twenty minutes, Officer Johnson encountered Mr. Johnson about “two blocks” away in an alley behind the 3700 block of Minnesota Avenue. Officer Johnson testified that “I didn’t know it was him right away,” but he stopped Mr. Johnson because Mr. Johnson was “the only individual in the area with . . . clothing” that “matched the lookout.” Officer Johnson testified that he and his partner handcuffed Mr. Johnson, elicited an admission from him that he had a gun in his bag for which he did not have a permit, and placed him under arrest. After Officer Johnson completed his testimony, the defense challenged the legality of the stop based on the informant’s tip, stressing that Officer Johnson had no information about the reliability of the informant whose tip had been relayed to him by Officer Cyphers, and no information about when the photograph that Officer Cyphers sent to Officer Johnson had been taken. Before hearing from the government, the court indicated that it believed the government had presented adequate evidence to demonstrate the legality of the stop; the court reasoned that Officer Cyphers’s knowledge about the reliability of the informant was “imputed” to Officer Johnson and that the fact that Mr. Johnson was located “at this particular  7 location wearing these articles of clothing” corroborated the tip. 2 The government then argued that the police had had reasonable articulable suspicion to stop Mr. Johnson based on the confidential informant’s tip, asserting that the informant had provided “particularized details” regarding the location and clothing of the person with the gun which were “independently corroborate[d]” by the screenshot of surveillance footage and by Mr. Johnson’s presence “in that location” only “about [twenty] minutes after.” The trial court “endorse[d]” the government’s argument, ruled “that there was a basis to properly stop [Mr. Johnson],” and denied Mr. Johnson’s motion to suppress. A jury subsequently found Mr. Johnson guilty of unlawful possession of a firearm, D.C. Code § 22-4503(a)(1); carrying a pistol without a license, id. § 22-4504(a); possession of an unregistered firearm, id. § 7-2502.01(a); and unlawful possession of ammunition, id. § 7-2506.01(a). This appeal followed. 2 Prior to this exchange with defense counsel, the court observed that the informant’s description of the man with a gun conveyed by Officer Cyphers to Officer Johnson was “identical” to the man in the screenshot who had been identified as Mr. Johnson by Officer Johnson. But after defense counsel observed both that informant had only given a description of clothing, and that Officer Cyphers had not told Officer Johnson when the screenshot was taken, the trial court asked whether the screenshot was “even necessary,” and did not mention it again.  8 II. Analysis In an appeal from a suppression ruling, this court “generally defer[s] to the trial court’s findings of fact unless they are clearly erroneous.” Mayo v. United States, 315 A.3d 606, 616 (D.C. 2024) (en banc). In this case, however, we have no factual findings to which to defer because the court simply “endorse[d] [the government’s] argument” at the suppression hearing and ruled the police lawfully stopped Mr. Johnson. But see Super. Ct. Crim. R. 12(d) (“When factual issues are involved in deciding a motion, the court must state its essential findings on the record.”). We thus conduct our own review of the facts, viewing them in the light most favorable to the suppression ruling, 3 and assess that ruling de novo. Mayo, 315 A.3d at 617 (rejecting any description of our authority to review the constitutional question as “limited” or “narrow in scope” or constricted to confirming that the trial court had a “substantial basis” for its ruling). The framework for our legal analysis is well established. Pursuant to Terry v. Ohio, 392 U.S. 1 (1968), the police may, consistent with the Fourth Amendment, briefly stop a person for investigative purposes, but only so long as the police possess 3 But see Mayo, 315 A.3d at 617 (acknowledging amicus’s challenge to this “light most favorable” standard on the ground that it is an anachronism, predating the Rule 12(d) requirement that trial courts make factual findings to support their rulings but declining to reach this issue because it was raised belatedly in a reply brief).  9 reasonable articulable suspicion of criminal activity. 4 “Although reasonable articulable suspicion requires considerably less than proof of wrongdoing by a preponderance of the evidence, officers must articulate more than an inchoate and unparticularized suspicion or hunch.” D.W. v. United States, 359 A.3d 166, 171 (D.C. 2026) (en banc) (internal citations omitted). “To determine if a Terry stop was supported by reasonable articulable suspicion, a court must examine whether the totality of the facts available to the officer at the moment of the seizure . . . warrant a [person] of reasonable caution in the belief that [the stop] was appropriate.” Mayo, 315 A.3d at 620 (internal quotation marks omitted). When the constitutionality of a stop is challenged, the government bears the burden of proving that the police had the requisite reasonable articulable suspicion. Griffin v. United States, 352 A.3d 1233, 1237 (D.C. 2026). The court in turn must do its job. “[A] court may not simply rely on a police officer’s conclusory assertions 4 Given that gun possession is a constitutionally protected activity, see U.S. CONST. amend. II; see also District of Columbia v. Heller, 554 U.S. 570 (2008); N.Y. State Rifle & Pistol Ass’n v. Bruen, 597 U.S. 1 (2022); United States v. Hemani, 146 S. Ct. 1677 (2026), “the mere fact that a suspect is in possession of a handgun, without more, is not sufficient to provide reasonable, articulable suspicion” to justify an investigative stop. Quarles v. United States, No. 25-CF-0019, slip op. at 7 (D.C. August 27, 2026); see also id. at 8-9; accord Doby v. United States, 359 A.3d 875, 881 (D.C. 2026) (acknowledging that “the mere possession of a firearm does not amount to reasonable articulable suspicion of wrongdoing”). Mr. Johnson, however, has made no argument that the police lacked reasonable articulable suspicion to stop him because they had insufficient basis to believe his possession of a gun was criminal.  10 in deciding whether a search or seizure was justified under the Fourth Amendment”; instead, it is the court’s responsibility to “evaluate the facts underlying those assertions.” Sanders v. United States, 751 A.2d 952, 955 (D.C. 2000); accord Parsons v. United States, 15 A.3d 276, 280 (D.C. 2011) (quoting Sanders); see also In re T.L.L., 729 A.2d 334, 341 (D.C. 1999) (recognizing that information in the possession of law enforcement “can contribute to the articulable suspicion calculus only if the judge has been apprised of sufficient facts to enable him to evaluate the nature and reliability of that information”). After all, “the scheme of the Fourth Amendment becomes meaningful only when . . . the conduct of those charged with enforcing the laws [is] subjected to the more detached, neutral scrutiny of a judge who must evaluate the reasonableness of a particular search or seizure in light of the particular circumstances.” Terry, 392 U.S. at 21. More specifically, when the police rely on a report of illegal activity to justify a seizure under the Fourth Amendment, as opposed to their own observations or knowledge, the court must determine whether the police reasonably relied on that report. Such reports “come in many shapes and sizes and from many different types of persons.” Illinois v. Gates, 462 U.S. 213, 232 (1983). Across the board, in conducting the requisite totality-of-the-circumstances analysis, courts look to facts regarding the identity of the person who made the report and their track record for supplying reliable information, to the extent it can be discerned; the informant’s  11 basis of knowledge; as well as to other circumstantial evidence that might lend support to the report. See, e.g., In re S.B., 44 A.3d 948, 952 (D.C. 2012) (conducting this analysis in an in-person report case); Joseph v. United States, 926 A.2d 1156, 1158-59 (D.C. 2007) (same in an identified 911-caller case); Barrie v. United States, 887 A.2d 29, 31-33 (D.C. 2005) (same in a paid informant case); Sanders, 751 A.2d 952 (same in case where an informant was “known” by voice only); Brown v. United States, 590 A.2d 1008, 1014 (D.C. 1991) (same in an anonymous informant case). We turn to examine these indicia of reliability here. We examine them separately before we “balance[] [our] assessment of the relative weights of all the various indicia of reliability (and unreliability) attending an informant’s tip,” Gates, 462 U.S. at 234, recognizing that “if one of the relevant indicia is deficient it is not fatal to a finding of [reasonable articulable suspicion] if there is a strong showing as to another, or if there exist some other indicia of reliability,” Goldston v. United States, 562 A.2d 96, 98 (D.C. 1989) (citing Gates); see also Porter v. United States, 7 A.3d 1021, 1024 (D.C. 2010) (citing Gates); accord Mayo, 315 A.3d at 621 (proceeding by “first assess[ing] the legitimacy and weight of each of the factors that possibly contribute to th[e] totality [of the circumstances analysis]” and “then weigh[ing] that information all together”). As noted above, we review the record facts in the light most favorable to the trial court’s ruling to make our own independent assessment whether a seizure was lawful, but this requirement that we  12 uphold a suppression ruling if “there is ‘any reasonable view of the evidence’ to support [it],” Mayo, 315 A.3d at 617 (quoting Ward v. United States, 365 A.2d 378, 381 n.3 (D.C. 1976)), does not compel us to draw unsupported inferences, fill evidentiary gaps, or rubber stamp a trial court ruling that is based on vague and conclusory testimony. A. The Informant’s Identity and Track Record The government presented limited evidence about the confidential informant who provided the tip relayed by Officer Cyphers to Officer Johnson. As Officer Johnson testified, the confidential informant was “known” to Officer Cyphers; Officer Cyphers “had been using [the informant] for a while”; and Officer Cyphers deemed the informant “reliable.” We begin with the testimony that the informant was known to the police. Standing alone, being “known” is not inherently informative. Our case law identifies three reasons this fact might matter, but each requires additional information to permit a reliability assessment. First, if an informant is known, their “reputation can be assessed.” Florida v. J.L., 529 U.S. 266, 270 (2000). But while the fact that an informant is “known” to the police may help the police gather information about their track record—a subject we discuss below—the mere fact that they are known does not assure reliability. A known informant could just as easily be a known liar.  13 Second, if an informant is known, they can “be held responsible if [their] allegations turn out to be fabricated.” Id.; see also Barrie, 887 A.2d at 32 (citing D.C.’s false- reporting statute and explaining that “the fact that the informant was well-known to the police and thus could be held to account were his information found to be willfully false, gives his report added weight”); Brown, 590 A.2d at 1016 (“A person who . . . voluntarily comes forward and identifies himself or herself, is more likely to be telling the truth because he or she is presumably aware of the possibility of being arrested for making a false report.”). But this concept of “knowing” assumes an ability to locate the informant via a name or an address. See, e.g., Nixon v. United States, 870 A.2d 100, 104 n.5 (D.C. 2005) (anonymous informant locatable because they reported a crime on their block to the police in-person); Groves v. United States, 504 A.2d 602, 605 (D.C. 1986) (911 caller locatable because they identified themself by name). Third, if an informant is known, their ability to serve as a reliable source of information and their incentives to lie can be assessed. See Joseph, 926 A.2d at 1164 (acknowledging that an informant’s reliability may be compromised “because of [a] possible bias or motive to falsify”); Goldston, 562 A.2d at 99 (detailing other considerations relevant to reliability including “personal attributes favoring accuracy in observation and reporting,” “personal connection with the suspect,” and “any circumstances suggesting probable lack of motivation to falsify”); see, e.g., Fleming v. United States, 923 A.2d 830, 835 (D.C. 2007) (including in the reliability calculus  14 the fact that, because the informant was “unpaid” and “did not use drugs,” they “had no reason to lie to gain monetary payment” or “to obtain drugs”); Barrie, 887 A.2d at 32 (including in the reliability calculus the informant’s “independence from drugs and alcohol,” and “current ability to remain out of the criminal justice system”). But again, for this rationale to do any work in a reliability calculus, the police must “know” the informant in more than a superficial way. Certainly, any assessment of the informant’s criminal history requires that the police possess sufficient information to locate the informant in public and official records. See, e.g., Fleming, 923 A.2d at 835 (noting that police were aware that informant had a pending criminal matter); Barrie, 887 A.2d at 31 (noting the police’s ability to determine that the informant was not on probation or parole and had not had any criminal convictions in decades). The government presented no evidence about how the informant was “known” to the police in Mr. Johnson’s case. Officer Johnson could not say whether Officer Cyphers knew the informant’s name, or phone number, or where he lived. 5 The record does not even establish whether Officer Cyphers spoke to the informant in-person or via phone or text or email. See Brown, 590 A.2d at 1016 (“[A]nonymity 5 We understand that there might be concerns about the officers disclosing this information in open court; our concern here is that Officer Johnson did not even represent that he or Officer Cyphers possessed this information.  15 takes on even greater significance where there has not even been a face-to-face confrontation between the person giving the information and the police.”) (alteration in original). We do not know if the informant was paid, which, as an expenditure of public funds, the government would presumably document. Thus, for all the government showed, Officer Cyphers “knew” the informant simply as someone he randomly encountered on the street, or could identify only by voice on the phone, or corresponded with via anonymous electronic communications such as a texting platform or a messaging app. See, e.g., Sanders, 751 A.2d at 952 (informant known to officer only by voice). This immediately weakens the value of the bare testimony that the informant was “known” to Officer Cyphers. Even if an informant is not sufficiently well known to allow an officer to locate them or to assess their incentives, an informant may have a “reputation” or a “track record” for giving reliable information. There is always a danger that individuals who regularly supply inculpating information to law enforcement on a confidential basis do so not to promote public safety but to benefit themselves or advance their own schemes. See Goldston, 562 A.2d at 99 (explaining “a paid police informant may generally be presumed to be less credible than a citizen informant”); cf. United States v. Bernal-Obeso, 989 F.2d 331, 333 (9th Cir. 1993) (“By definition, criminal informants [seeking reward for their testimony] are cut from untrustworthy cloth and must be . . . carefully watched by the government and the courts.”). Thus,  16 a track record—that is, data about “the number, frequency, content, accuracy . . . and productivity of any past tips”—is “[p]erhaps the most telling indic[ator] of reliability.” Sanders, 751 A.2d at 955; accord Porter, 7 A.3d at 1024 (quoting Sanders, 751 A.2d at 955); see also Parsons, 15 A.3d at 280 (recognizing “an informant’s history of supplying prior productive information is a most important guide to establishing reliability and credibility,” quoting United States v. Boxley, 985 A.2d 1108, 1112 (D.C. 2009), and citing cases “reflect[ing] th[is] importance”). 6 We cannot say here that the government even “thinly developed” the confidential informant’s track record, a descriptor we applied in Sanders, 751 A.2d at 955. In that case, the officer who received the tip testified that, in the “five or six times” they had spoken, “the tipster had never given [him] incorrect information.” Id. We held that this showing “might have been enough to create articulable suspicion for a Terry stop,” but did not suffice to show probable cause. Id. at 955-56 (noting inter alia that the officer had not detailed “how many tips were[, like the tip that resulted in the seizure at issue,] drug-related, how many tips were eyewitness, or over how long a period the tips were made”). We have far less information about 6 Although Mr. Johnson highlights that he made discovery requests for this sort of information, he cites no law to support the contention that the trial court in this case abused its discretion in denying his requests. See Jones v. United States, 317 A.3d 373, 380-81 (D.C. 2024). Thus, we do not understand him to raise this as a separate argument.  17 the informant’s track record in this case than in Sanders. We have no idea how many times Officer Cyphers had previously communicated with or received information from this informant, or how recently. And although we know he “had been using [the informant] for a while,” we know neither what it means to “use” the informant, nor how long “a while” is in Officer Cyphers’ estimation. This leaves the assessment, relayed by Officer Cyphers to Officer Johnson and by Officer Johnson to the court, that the informant was “reliable.” But this assessment is entirely conclusory and communicates no facts to the court to allow it to probe that assessment, as our case law requires. See Parsons, 15 A.3d at 280 (prohibiting a trial court from “simply rely[ing] on a police officer’s conclusory assertions” and requiring instead that the court “evaluate the facts underlying those assertions”) (quoting Sanders, 751 A.2d at 955). The trial court attempted to cure the deficiency in the government’s evidence by imputing Officer Cyphers’s basis of knowledge to Officer Johnson, see supra Part I, but, as the government implicitly concedes by declining to defend this reasoning, it erred in doing so. Under the collective knowledge doctrine, a trial court may aggregate the information shown to be collectively possessed by officers involved in a seizure. See Milline v. United States, 856 A.2d 616, 620 (D.C. 2004); McFerguson v. United States, 770 A.2d 66, 72 (D.C. 2001). Where an officer is not  18 called to testify about what he knew, however, the trial court may not simply assume that the non-testifying officer possessed a sufficient factual basis for a conclusory statement repeated by a testifying officer. We made this clear in Parsons. There, just as here, the officer who acted on the tip had never worked with the informant, and the officer who had and “presumably [knew of] the informant’s track record and the existence of other virtuous qualities” did not testify; thus, “even applying the collective knowledge doctrine, there was no testimony from which the trial court could have judged the informant’s credibility” via an examination of past performance. 7 Parsons, 15 A.3d at 280-81; cf. Jenkins v. United States, 152 A.3d 585, 589-91 (D.C. 2017) (holding that the government had failed to present evidence showing a defendant’s seizure was based on reasonable articulable suspicion where 7 The government seeks to distinguish Parsons on the ground that it “involved a determination of whether the police had probable cause to search a person” for contraband “based on a confidential informant’s tip, not reasonable articulable suspicion,” which “is a less demanding standard.” But regardless of the quantum of proof required, the same totality of the circumstances analysis is conducted when examining the reliability of an informant’s tip. See Alabama v. White, 496 U.S. 325, 330-31 (1990) (explaining that “[t]he same [totality of the circumstances] approach applies in the reasonable suspicion context, the only difference being the level of suspicion that must be established”). Where this court has already held in Parsons— a case evaluating probable cause on a comparably anemic record—that the trial court had no basis to assess an informant’s track record, it is of no moment that, “in applying” this same analysis, “allowance must be made . . . for the lesser showing required to meet [the reasonable articulable suspicion] standard,” 496 U.S. at 328-29. The government could not clear even a lower bar without presenting the trial court some nonconclusory evidence with which to independently assess the informant’s track record.  19 the officer who effected the seizure relied on a description from an officer who had watched surveillance video footage but the government did not call that officer to testify about what he had seen in that footage or move that footage into evidence). Were it otherwise, officers could “bring about a lawful stop by the simple expedient of passing information [from one officer who does not testify] on to another officer,” who does not know what the first officer does or does not know. Jenkins, 152 A.3d at 590 (rejecting such a rule) (citation modified); see also In re E.A., 343 A.3d 1, 6-7 (D.C. 2025) (quoting Jenkins and recognizing that “the information conveyed over a radio dispatch can contribute to the articulable suspicion calculus only” if the trial court has an independent basis to evaluate the “nature and reliability” of the information in that dispatch) (citation modified). But that would hollow out the judiciary’s role in overseeing searches and seizures. In sum, the trial court was given no information about the identity or track record of the confidential informant that would have permitted the court to independently assess the informant’s reliability. This was largely due to the fact that the officer who received the tip from the informant did not testify. Indeed, we are unaware of any case in which this court has determined that the evidence presented about a confidential informant’s identity and track record supported a meaningful assessment of reliability where the government did not call to the stand the officer who received the tip.  20 B. Other Indicia of Reliability Our inquiry does not end just because the trial court could not meaningfully assess the informant’s reliability based on their identity or track record. As noted above, a strong showing as to other indicia of reliability may fill the evidentiary gap. The analysis is always fact-specific, but precedent provides some guide as to what facts may supply the necessary objective reassurance of reliability. Courts have recognized, for example, that an informant’s ability to accurately predict future events can bolster reliability by indicating access to inside information. 8 See, e.g., White, 496 U.S. at 331 (anonymous tip deemed reliable where informant accurately predicted defendant would leave a particular apartment at a particular time, get into a particular vehicle and drive to a particular motel); Gates, 462 U.S. at 244-45 (anonymous tip deemed reliable where informant accurately predicted that defendant would fly to Florida “in the next day or so” and then drive his family car back north loaded with drugs); see also Brown, 590 A.2d at 1016 (explaining an otherwise weak tip such as “[w]here the informant . . . provides information as to facts or events observable by anyone 8 Of course, the fact that the informant accurately predicts that an individual possessed contraband does not count. We consider only the information that the police possessed before they seized a defendant. Brown, 590 A.2d at 1013 (“A search is not to be made legal by what it turns up; it is good or bad when it starts and does not change character from its success.”).  21 (e.g., that a man in a white T-shirt and blue jeans is standing at a given corner)” could become reasonable articulable suspicion with predictive information). The government presented no such predictive information from the informant in this case. It argues instead that it presented direct evidence of the informant’s basis of knowledge: testimony from Officer Johnson that Officer Cyphers told him that the informant “saw” the man with the gun. The government refers to Officer Johnson’s trial testimony on cross-examination when in response to defense counsel’s question, “[s]o you don’t know how the confidential informant saw the weapon?” Officer Johnson responded, “No, I don’t have any knowledge of their conversation[,] just that he saw him with one, so.” Even considering testimony that (1) was not presented at the suppression hearing and thus could not have been a foundation for the ruling we are reviewing, and (2) incorporated an assumption of defense counsel who had no information  22 about what the informant knew, 9 this bare assertion by the informant without any detail of how or when this “seeing” occurred is not much in the way of eyewitness evidence. Even coupled with Officer Johnson’s suppression hearing testimony, all we know is that the informant said he saw a man with “a firearm” of some unspecified make, model, or color, somewhere “on his person.” This compares poorly to other cases cited by the government where courts have held that the details given by an informant-eyewitness supported the reliability of a tip. See, e.g., Navarette v. California, 572 U.S. 393, 399-400, 404 (2014) (acknowledging that it was a “close case” but that the 911 caller’s report that a silver Ford F-150 pickup with license plate 8D94925 had run her off the road indicated that the caller was an eyewitness to reckless driving and gave the police reasonable articulable suspicion 9 The government cites this court’s decision in Mayo for the proposition that, in evaluating the trial court’s suppression ruling, this court may look to “the evidence presented at the suppression hearing and undisputed evidence at trial.” But in Mayo we acknowledged that our case law regarding the use of trial testimony in evaluating a suppression ruling was inconsistent. 315 A.3d at 623 n.8; see also id. at 641 (McLeese, J., dissenting) (doubting the “wisdom and fairness” of permitting consideration of such testimony to either support or undermine a suppression ruling). Thus, we merely “assumed” without deciding that, even if the “use of undisputed trial testimony is proper,” the testimony highlighted by the government was undeserving of “much . . . weight.” Id. at 623 n.8. Nevertheless, because Mr. Johnson cited to the trial testimony in his discussion of the suppression ruling and did not file a reply brief contesting the government’s reliance on the one strand of trial testimony that diverged from Officer Johnson’s testimony at the suppression hearing, we consider forfeited any argument that we should refrain from considering Officer Johnson’s trial testimony.  23 to conduct the traffic stop 10); (Melvin) Jackson v. United States, 109 A.3d 1105, 1106-08 (D.C. 2015) (holding police had reasonable articulable suspicion to seize defendant based on an anonymous 911 caller’s report that she had seen the individual she described “bring out a silver pistol out of his pocket,” that it “scared the hell out of [her],” and that she then called the police); Boxley, 985 A.2d at 1100, 1113 (holding police had reasonable articulable suspicion where paid informant who had helped with eleven prior drug arrests and had never given unreliable information told police that he had seen the defendant selling drugs, specifically crack cocaine). We take Officer Johnson’s trial testimony for what it is worth. In addition, the government argues that this is a case where “innocent details provided by the informant were verified.” We acknowledge that, as corroboration of reliability, our court has looked to the ability of the police to swiftly corroborate innocent details of the informant’s report. See e.g., Porter, 7 A.3d at 1024 (explaining that the reliability of a paid informant who had worked with the police 10 The Court in Navarette also acknowledged that reliability of a citizen informant’s tip may be bolstered by the existence of systemic safeguards—like the fact that 911 calls are recorded and callers can be traced even if they do not provide identifying information—that might deter false reporting. 572 U.S. at 397; but see Miles v. United States, 181 A.3d 633, 639 n. 12 (D.C. 2018) (noting the “affirmative evidence in the record that the 911 system here did not work in the way that gave the . . . Court [in Navarette] confidence in the 911 system’s ability to discourage dishonest tipsters”). No such safeguards are present in this case.  24 for ten years, never given false information, and facilitated 100 arrests was “bolstered” where “[w]ithin two minutes from receiving the tip” officers found man where informant had said he would be matching the informant’s specific description); Boxley, 985 A.2d at 1110, 1112 (holding that reliability of a paid informant who had helped the police make eleven arrests and never provided false information was corroborated where police discovered man matching informant’s description at the location the informant had said he saw the man selling drugs); see also Joseph, 926 A.2d at 1163 (holding that “the identified [citizen] informant’s telephone tip [via 911] and the officer’s immediate corroboration [in less than a minute] of [suspect’s] location and clothing were sufficient to establish a reasonable articulable suspicion”). But as these cases reflect, such innocent details did not form the foundation for reasonable articulable suspicion. Rather, the “reasonable suspicion” needed to justify a Terry stop “requires that a tip be reliable in its assertion of illegality, not just in its tendency to identify a determinate person,” and merely corroborating innocent details—such as a person’s location and clothing—does not establish that the tip is reliable regarding alleged criminal activity. J.L., 529 U.S. at 272; see also Miles, 181 A.3d at 638 (quoting J.L. and explaining that, in cases where police receive an anonymous tip alleging the person identified or described had a gun, the police “must typically see something that confirms the presence of a gun”) (citation modified); accord Brown, 590 A.2d at 1016-17 (explaining that where an  25 informant who “successfully predicts events that have not yet occurred at the time of the tip but which are thereafter verified by police” and relays “facts . . . observable by anyone” that are corroborated by the police, the police may possess “sufficient indicia of the informant’s credibility (as well as his or her basis of knowledge) to justify a stop”). In any event, the government’s “corroboration-by-innocent-details” evidence in this case is weak because it was vague in multiple ways. First, the tipster’s description of the man with the gun was generic. The tipster did not provide the height, weight, race, age, or hairstyle of this