Practice in a safe space, then meet the real thing
Across the arc of human difference and the structures built to maintain order, a quiet collision has long produced outsized harm — autistic individuals encountering police in moments neither party is prepared for. Researchers in Philadelphia have now tested whether virtual reality can serve as a bridge, offering autistic teens and adults a safe space to rehearse high-stakes encounters before they happen in the world. A clinical trial found that those who trained in a VR simulation fidgeted less and behaved more adaptively during live interactions with real officers afterward — small but measurable signs that preparation can soften the edges of a dangerous gap.
- Autistic individuals face a well-documented elevated risk of harm during police encounters, where difficulty reading social cues can turn a routine stop into a crisis.
- Families and autistic adults have long carried a quiet, reasonable fear about these interactions — a fear that researchers at CHOP and St. Joseph's University decided to take seriously as a clinical problem worth solving.
- The team built a mobile virtual reality module allowing participants to rehearse police encounters repeatedly and safely, then tested it against a video-based alternative across 47 participants aged 12 to 60.
- After just three sessions, VR-trained participants showed real, observable improvements during live encounters with Philadelphia police officers — less fidgeting, better overall responsiveness.
- The study stands as one of the rare clinical trials focused on life-skills outcomes for autistic adults, and its findings point toward a scalable tool that could reduce danger and anxiety for individuals and families alike.
Autistic people encounter police more often than their non-autistic peers, and when they do, the stakes are high. The social and sensory demands of an unexpected encounter with a uniformed officer can quickly spiral into misunderstanding — and the accumulation of stories about violence against autistic and disabled individuals during police stops has made fear of these moments a reasonable part of daily life for autistic adults and those who love them.
Researchers at Children's Hospital of Philadelphia and St. Joseph's University asked whether virtual reality could help. They built the Floreo Police Safety Module — a mobile VR intervention designed to let people rehearse police encounters in a controlled, repeatable setting — and enrolled 47 autistic individuals between the ages of 12 and 60. Half used the VR module; half watched a video-based intervention called BeSAFE The Movie. Both groups completed three 45-minute sessions over roughly nine days. Then came the real test: a live interaction with an actual Philadelphia police officer.
Both groups retained skills from their training, but the VR group showed something distinct. During the live encounters, they fidgeted less than before and demonstrated measurably better overall behavior and responsiveness. These were not dramatic transformations — but they were real, observable changes during an interaction that many autistic individuals find deeply threatening.
Senior researcher Julia Parish-Morris emphasized the importance of measuring outcomes against actual police encounters rather than simulations alone. First author Joseph McCleery credited the Philadelphia Police Department, whose officers volunteered their time across weeks of study participation to make that real-world measurement possible.
The broader implication is significant: if three brief sessions of VR training can meaningfully shift how autistic individuals present themselves during high-stakes encounters, the potential to scale this preparation is real. For autistic teens and adults navigating the transition to independence — and for the families who worry about what a police stop might mean — evidence-based tools like this one represent something more than a clinical finding. They represent a measure of safety that did not exist before.
Autistic people encounter police at higher rates than their non-autistic peers, and when they do, the interaction carries real danger. The difficulty reading social cues that characterizes autism, combined with the stress of an unexpected encounter with a uniformed officer, can spiral quickly into misunderstanding and harm. Stories of violence against disabled and autistic individuals during police stops have accumulated in news reports and academic literature long enough that fear of these encounters has become a reasonable part of life for autistic adults and the people who care about them.
A team of researchers at Children's Hospital of Philadelphia and St. Joseph's University decided to test whether virtual reality could help. The logic was straightforward: autistic individuals need practice navigating police interactions, but arranging actual encounters with real officers for training purposes would be expensive, time-consuming, and impractical. A simulation could offer the repetition and safety that real-world practice could not. They built the Floreo Police Safety Module, a mobile virtual reality intervention designed to let people rehearse these encounters in a controlled setting.
To test whether the simulation actually worked, the researchers enrolled 47 autistic individuals ranging in age from 12 to 60 years old. Half were assigned to use the virtual reality module; the other half watched BeSAFE The Movie, a video-based intervention. Both groups completed three 45-minute sessions spread across nine days on average. The crucial part came next: after training, each participant had a live interaction with an actual Philadelphia police officer. The researchers measured what happened.
The results showed that both groups retained and used skills they had learned. But the virtual reality group showed something the video group did not. When they met with the real officers afterward, the VR-trained participants fidgeted less than they had before the training. Their overall behavior and responsiveness during the live encounters improved measurably. These were not dramatic transformations, but they were real, observable changes in how autistic individuals presented themselves during an interaction that many find deeply stressful.
Julia Parish-Morris, the senior researcher and a psychologist at CHOP's Center for Autism Research, emphasized the importance of testing the intervention against actual police encounters rather than relying on self-report or simulated measures alone. "Using mobile virtual reality is a great way to engage people and provide a flexible, personalized opportunity to practice specific skills," she said. "But we also felt it was very important to test the efficacy of our trial using a practical and generalizable outcome measure, which was a live interaction with a real police officer."
The study, published in the Journal of Autism and Developmental Disorders, represents one of the few clinical trials to enroll autistic adults with the explicit goal of improving life skills after the transition to adulthood. Joseph McCleery, the first author and a psychologist at St. Joseph's University, credited the Philadelphia Police Department for their cooperation. Multiple officers volunteered their time over weeks and months to participate in the study, allowing researchers to measure whether the virtual reality training would translate into real-world benefit.
The implications reach beyond the immediate participants. Autistic individuals face a documented elevated risk of negative outcomes during police encounters—outcomes that can range from unnecessary escalation to serious injury. If a brief intervention, delivered across just three sessions, can measurably improve how autistic people present themselves and respond during these high-stakes interactions, the potential to scale this training becomes significant. The work suggests that targeted, evidence-based preparation could help autistic teens and adults navigate the transition to independence with greater safety and less anxiety, while also reducing the fear that families carry about what might happen if their autistic loved one is ever stopped by police.
Citações Notáveis
We felt it was very important to test the efficacy using a practical and generalizable outcome measure, which was a live interaction with a real police officer.— Julia Parish-Morris, senior study author and psychologist at CHOP
Multiple police officers dedicated their personal time to allow us to determine whether the virtual reality intervention would truly help autistic people feel and perform better during live police interactions.— Joseph McCleery, first author and psychologist at St. Joseph's University
A Conversa do Hearth Outra perspectiva sobre a história
Why does this matter so much? Autistic people interact with police—sure, but is it really a widespread problem?
The data shows autistic individuals have higher rates of police contact than non-autistic peers. And when those encounters happen, the stakes are different. An autistic person might not read a officer's tone correctly, or might not respond in the way an officer expects, and that gap in communication can escalate quickly.
So the virtual reality training teaches them what to do and say?
It's more subtle than a script. It's practice in a safe space. They can experience the sensory and social elements of a police encounter—the uniform, the authority, the unexpected nature of it—without real consequences. Then they can try again.
And the study showed it actually worked?
The group that trained in VR showed measurable changes when they met with real officers afterward. They fidgeted less, responded better. The control group that watched videos didn't show those same improvements.
Why is fidgeting the thing that changed?
Fidgeting is often how anxiety shows up in autistic people. If they're less fidgety, it suggests they're calmer, more regulated. An officer might perceive that as cooperation or clarity, rather than misinterpreting it as evasiveness or disrespect.
So this could actually save lives?
It could prevent escalation. It could give autistic people tools to communicate more effectively in a moment when miscommunication carries real risk. That's not nothing.