Imagine police officers in Charlotte, North Carolina, are following a car that they find suspicious. The male driver parks and the officers follow him as he joins a group of five other men, who are standing in a circle, talking and laughing. One of the officers identifies a member of the group as someone he knows has a prior drug arrest. Police then approach the group, and one friend tells the officers that he is legally carrying a firearm. The police decide to frisk the entire group, ultimately finding an illegally possessed firearm on another friend, who challenges the charges brought against him in court. During that challenge, the police officer testifies that he has been trained to suspect that when one person in a group is carrying a weapon, as least one other is as well. Meanwhile, the individual insists that police officers relied on associational suspicion in violation of his Fourth Amendment rights.

This encounter, based on a recent case, reveals the harsh reality that the Fourth Amendment’s reasonable suspicion standard—what the Supreme Court says is required for police to engage in a brief, investigatory stop of individuals—is in jeopardy. Increasingly, police are relying on vague, racialized characterizations of human behavior and judges are deferring to police. My research has shown that this occurs in a cyclical pattern: police increasingly cite a new factor justifying intervention, courts defer to that factor to affirm the interventions, scholars and some courts recognize the factor’s problematic features, it’s usage slowly wanes, and then another insidious term takes its place.

I’ve been studying police reliance on a person’s suspicious companions as justification for intervention—stops, searches, pat downs. In several contexts of discretionary, proactive policing, police rely on the behavior of a person’s companions to justify their intervention. For example, a companion’s consent to an officer’s request—even when it is not legally required—makes an individual’s refusal to comply appear suspicious in comparison. Or an officer’s criminal suspicion regarding one person may be valid but is inappropriately deflected onto all other companions. Such expanded suspicion is prominent in pedestrian stops, such as the North Carolina example described. Police approach a group, and even if articulable suspicion exists for one friend, it is unconstitutionally deflected onto each of their companions.

Police also rely on a person’s suspicious companions when those friends are not even present for the encounter. For example, police rely on flawed and overly broad gang designation lists in a manner that undermines the individuality of suspicion required for a police stop or search. Finally, when individuals associate with any of the nearly five million Americans serving a heavily surveilled community supervision—probation, supervised release, and alternatives to incarceration—these individuals are subject to criminal suspicion because of their social associations.

Reliance on associational suspicion is antithetical to Fourth Amendment doctrine. The leading Supreme Court precedents that define the level of criminal suspicion required for police to conduct an intervention emphasize individuality of suspicion. Each of the two primary standards of criminal suspicion, probable cause and reasonable suspicion, are consistently interpreted to require suspicion directed towards an individual. And where the Court has addressed police treatment of suspicious groups, or exceptions for reducing privacy rights, it presumes the primacy of individualized exceptions.

Police reliance on associational suspicion, separate and apart from being unsupported by legal precedent, is troubling for other reasons. There is significant racial bias in the degree to which an individual is perceived by police to be part of a group versus existing on their own. Recent empirical work reveals that decision makers are more likely to infer group liability in cases involving defendants of color, yet more likely to treat white defendants as individuals. Survey participants were more likely to quickly group together Black and Latino names with words associated with groups, and white faces with individuality. Given these racialized assumptions about social connection, the Fourth Amendment’s erosion will occur unequally.

Finally, in our era of increasingly legal gun ownership, suspected gun possession becomes presumptively lawful and will no longer serve as a legitimate basis for police to suspect criminal activity. But given what we know about racialized discretionary policing, where suspected gun possession can no longer legally justify a police officer’s decision to stop or search, reliance on a person’s suspicious companions, can easily fill that gap.

In all of these situations and more, the expansion of criminal suspicion to justify police intervention undermines our democracy — suspicion without “reasonable” is like a net without holes. It catches everyone.

March 2026