When a microphone on a city utility pole reports gunfire, and a police officer shows up moments later and finds someone standing nearby, can the officer stop that person? That is, roughly, the question at the heart of Commonwealth v. Foster, decided by the Pennsylvania Supreme Court on May 19, 2026. The Court said yes — the stop here was lawful — but the more interesting part of the decision is how it got there, and what four separate opinions reveal about a Court still working out how to handle gunshot-detection technology under our search-and-seizure law.
A quick word on the technology. ShotSpotter is an acoustic detection system used by the City of Pittsburgh. Sensors placed around the city pick up loud noises that might be gunfire. Those sounds get reviewed — partly by algorithm, partly by human analysts — to decide whether they really are shots, and the sensors’ locations are used to triangulate roughly where the noise came from. If the system concludes it heard gunfire, it sends an alert to the police. The whole process takes somewhere in the neighborhood of thirty to forty-five seconds.
The Stop
Around 2:00 a.m. on a September morning in 2019, Pittsburgh Police Officer Nathan Powers got a ShotSpotter alert: a single shot near 1439 Hoffman Street in the Manchester neighborhood. On his way there, a second alert came in — four more shots, same location. Officer Powers was a block or two away. He turned onto Hoffman Street, activated his emergency lights, and saw a car parked against the flow of traffic with its headlights on. Jamar Foster was in the driver’s seat; a woman named Tiffany Towns was in the passenger seat.
As Officer Powers approached, Foster got out of the car and walked toward a house. Towns, meanwhile, appeared to be “moving around in the car trying to grab things.” The officer stopped, got out, and ordered Foster to return to the street. Foster kept walking. Officers drew their guns, ordered him down at gunpoint, and handcuffed him on the ground. The investigation that followed turned up evidence that Foster was driving under the influence, and he was eventually convicted of DUI and driving with a suspended license.
Foster moved to suppress everything that came after the seizure, arguing the police had no lawful basis to stop him in the first place. The trial court denied the motion, the Superior Court affirmed, and the Supreme Court took the case to address a narrow question: did the Superior Court give too much weight to Foster’s closeness in time and place to the ShotSpotter alert when it found reasonable suspicion?
The Legal Frame: Three Kinds of Police Encounters
To understand why the timing of the stop matters so much, you need the basic architecture of search-and-seizure law. Pennsylvania courts sort police-citizen interactions into three categories. The lightest is a “mere encounter” — an officer walking up and asking a question — which requires no suspicion at all because you’re free to walk away. The heaviest is an arrest, which requires probable cause. In between sits the “investigative detention,” often called a Terry stop after the U.S. Supreme Court’s 1968 decision in Terry v. Ohio. A Terry stop is a real seizure — you are not free to leave — but it can be justified on less than probable cause. The officer needs “reasonable suspicion”: specific, articulable facts suggesting criminal activity is afoot.
Reasonable suspicion is not a demanding standard. The U.S. Supreme Court has said it requires far less than a preponderance of the evidence and “falls considerably short” of being more-likely-than-not. But it does have to be particularized — tied to the specific person being stopped, not just a vague sense that something is wrong somewhere.
Two features of the doctrine drive the whole case. First, a Terry stop must be justified at its inception. Whatever facts the officer develops after the seizure cannot be borrowed backward to rescue a stop that was unsupported when it began. Second, courts assess reasonable suspicion under the “totality of the circumstances” — no single fact has to carry the day, and even a combination of innocent-looking facts can add up to something.
So everything turns on two questions: when exactly was Foster seized, and what did the officer know at that precise moment?
When the Seizure Happened
The Court, in an opinion by Justice Dougherty, held that Foster was seized when Officer Powers ordered him back to the street. That conclusion rests on settled Pennsylvania law: when an officer commands a citizen to stop, a reasonable innocent person would not feel free to keep walking, so a seizure has occurred. (This is one of the places where Pennsylvania’s Article I, Section 8 is more protective than the federal Fourth Amendment — under federal law, a person who ignores a command to stop and keeps moving has not yet been “seized,” but under our state constitution the order itself does the work.)
There was a wrinkle worth flagging, because it surfaces again in the dissent. Foster had argued in the Superior Court that he was actually seized earlier — the moment Officer Powers rounded the corner with his lights on. The majority held that argument waived, both because Foster abandoned it on appeal to the Supreme Court and because he never raised it in the trial court. The majority also noted, almost in passing, that the officer never even ran his siren and never pulled alongside Foster’s car, which distinguishes this situation from the lights-and-positioning that produced a seizure in Commonwealth v. Livingstone.
What Counted, and What Didn’t
Here is where the opinion is most useful, because the majority did not simply bless every fact the Commonwealth offered. It threw one out.
The Commonwealth had leaned on the idea that this was a “high-crime area.” The majority would have none of it. The only evidence in the record was the officer agreeing, in response to a leading question, that the neighborhood had its “hot spots” and had been one “in the past.” The majority — drawing on its recent decision in Commonwealth v. Lewis, 343 A.3d 1016 (Pa. 2025) — held that this kind of testimony is too vague to prove anything. Lewis established that the Commonwealth has to show an area was high in crime at the time of the stop, and that “merely intoning buzzwords is never sufficient.” Officer Powers’s testimony was so substantively and temporally fuzzy — a “hot spot” at some undefined point in the past — that the majority concluded a finding of a high-crime area simply wasn’t supported by the record. That factor was out.
That left four things the majority did credit, and it was careful to say that none of them alone would be enough:
The two ShotSpotter alerts indicating five shots near a specific address, suggestive of gun crimes or the unlawful discharge of a firearm, at 2:00 a.m. in a residential neighborhood. The lateness of the hour mattered — it cut against innocent explanations like daytime target practice.
The fact that Foster and Towns were the only people present at the location. Arriving within seconds of the second alert, the officer wasn’t sorting through a crowd of possible suspects. He had two people, and a rapid response time that left little opportunity for a shooter to slip away. On this point the majority reached for out-of-state authority — State v. Nimmer from Wisconsin and United States v. Jones from the D.C. Circuit — both finding reasonable suspicion where an officer arrived quickly after a ShotSpotter alert and the defendant was the only person around.
Towns’s movements inside the car as the police approached — reaching around, grabbing for “things.” The majority treated this as relevant to suspicion of Foster, not just Towns, citing a very recent per curiam decision from the U.S. Supreme Court, District of Columbia v. R.W. (2026), for the proposition that a passenger’s suspicious conduct can cast the driver’s situation in a suspicious light.
And Foster’s own conduct — getting out of the car at the sight of the police and walking toward a house — which the majority characterized as “evasive.” Not flight, the majority was careful to say, and certainly not the “headlong flight” that featured in Illinois v. Wardlow. Just evasive. Avoidance of contact with the police, in response to the police, is a permissible factor in the mix.
Add it all up — late-night alerts of gunfire, the rapid response, the two of them alone at the scene, and the pair’s furtive and evasive behavior as police arrived — and the majority found reasonable suspicion particular to Foster. The stop stood, and the Superior Court was affirmed.
In my view, the most important thing the majority did was not the result but the housekeeping. By tossing the high-crime-area factor for lack of record support, the Court reinforced Lewis and put real teeth into the requirement that the Commonwealth actually prove this stuff rather than gesture at it. Defense lawyers should read those passages closely. The “high-crime area” label has done a lot of damage in suppression litigation for years, and Foster — even as a loss for the defendant — tightens the screws on it.
What the Concurrence Added
Justice Mundy joined the result but wrote separately to defend the very factor the majority discarded. In her view, Officer Powers’s testimony about the area’s “hot spots” was grounded in his years of personal experience patrolling that zone, which she thought enough to distinguish it from the bare “buzzwords” condemned in Lewis. She also pointed out that Foster himself had repeatedly described the location as a high-crime area in his own briefs — a concession she would have held him to. The majority answered both points in its footnotes: personal experience does not automatically make vague testimony specific, and a defendant reciting a phrase in a brief is not the kind of clear, unequivocal admission that binds him. The disagreement is narrow, but it’s a live one, and it tells you the Court is not unanimous on how much daylight Lewis really put between acceptable and unacceptable high-crime-area proof.
What the Dissents Added
Justice Donohue would have reversed. Her central complaint is that the majority dodged the question the Court actually granted review to decide. The Court took the case to address how much weight a ShotSpotter alert deserves in a reasonable-suspicion analysis — an issue about an emerging technology that is going to keep coming up. Instead, the majority piled up a handful of other factors and avoided the hard question. And the factors it piled up, in her telling, don’t survive contact with the record.
On Towns’s “furtive movements”: the officer never used the word “furtive,” never described conduct that fit the definition, and never said her movements raised his suspicion of Foster. Donohue traces how the “furtive” gloss entered the case — not from the officer’s testimony, but from the prosecutor’s argument afterward — and points out that neither the suppression judge nor the Superior Court relied on Towns’s behavior at all. Grabbing a purse before getting out of a car, she writes, is not stealthy, surreptitious, or sly. It’s what people do.
On Foster’s “evasive” behavior: the record showed he started walking away before the lights came on, doing exactly what you’d expect of someone who just parked in front of a house at 2:00 a.m. The only conduct that could fairly be called evasive — refusing to return to the street — happened after the order to stop. And conduct after the seizure can’t justify the seizure. That’s the inception rule, and Donohue would enforce it.
With those factors stripped away, she’s left with what she sees as the only honest basis for the stop: Foster’s proximity in time and space to the alert. ShotSpotter takes thirty to forty-five seconds just to classify a sound and send the alert, so by the time Officer Powers arrived, somewhere between forty and sixty seconds had passed since the shots — plenty of time, she notes, to leave the scene on foot or by car. And the location is only accurate to within eighty to one hundred feet; the officer himself admitted he couldn’t rule out that the shots came from a different block. In a dense residential neighborhood, that margin matters enormously. Standing alone, she concluded, spatial and temporal proximity to a ShotSpotter alert was not enough to tie Foster to the suspected crime.
Justice Wecht joined Donohue with one exception. He would go further than the majority on the high-crime-area issue — not just hold that the proof failed here, but reject the “high-crime area” as a legitimate factor in the reasonable-suspicion calculus altogether. His point is conceptual: the fact that police allocate patrols and resources to certain neighborhoods reflects real judgments about where crime occurs, but those judgments are about places, not people. They can’t substitute for the individualized suspicion the constitution requires. Investigatory tools, in his framing, are one thing; evidentiary requirements are another, and the Court should not let the first stand in for the second. ShotSpotter, he adds, falls in the same bucket — a useful tool that cannot do the work of individualized suspicion.
Where This Leaves Things
Foster is a narrow decision dressed up in a significant question. The Court affirmed a stop, declined to say much of anything about ShotSpotter’s reliability as technology, and reached its result by leaning on the surrounding circumstances rather than the alert itself. Practitioners hoping for a clear rule on how courts should treat gunshot-detection alerts will be disappointed — that question remains open, and the dissent’s framing of it is a roadmap for the next litigant who preserves the issue and builds a record on the technology’s limits.
But the decision is not nothing for the defense. The majority’s treatment of the high-crime-area factor, building on Lewis, is a real constraint on a label police and prosecutors have invoked loosely for years. And the dissents lay out, in careful detail, the arguments that a better-developed record might have won on: the gap between when shots are fired and when an alert lands, the slack in ShotSpotter’s locational accuracy, and the difference between conduct that precedes a seizure and conduct that follows it. The defendant who challenges ShotSpotter’s reliability head-on, preserves it cleanly, and puts on evidence about how the system actually works may get a very different answer. Foster didn’t close that door. It just declined to walk through it.