Autistic people gain measurable police-interaction skills after short VR training

Participants who used the Floreo Police Safety Module (PSM) showed reduced fidgeting, improved social responses, and better performance during live assessment interactions, consistent with reduced observable anxiety-linked behavior and improved police-interaction responses. Both intervention groups also reported greater comfort and confidence when interacting with police officers.

These findings highlight the potential of immersive tools to support safer interactions in high-pressure situations, where misunderstandings can quickly escalate.

Police interactions can quickly become stressful and dangerous, particularly for autistic individuals who may struggle with unfamiliar social cues, sensory overload, or rapid verbal instructions. Studies show that autistic people face a higher risk of negative outcomes, including injury, use of force, and discrimination during law enforcement encounters.

Although autism-focused police training programs are expanding, many departments in the United States still lack adequate preparation to support neurodiverse individuals. Existing interventions often rely on role-playing exercises, but these approaches are difficult to scale and rarely capture the unpredictability of police encounters, highlighting the need for accessible, practical training tools.

About the study

In this RCT, researchers evaluated whether Floreo PSM could help autistic individuals navigate police encounters more effectively than a video-based educational program, BeSAFE The Movie.

The study included 47 verbally fluent autistic participants aged 12 to 60 years who met Diagnostic and Statistical Manual of Mental Disorders, fifth edition (DSM-5) criteria for autism. Participants had an intelligence quotient (IQ) score of at least 75, as measured by the Wechsler Abbreviated Scale of Intelligence (WASI-II). They were randomly assigned to either the VR or video-based training group.

Before training, participants completed standardized autism assessments, including the Autism Diagnostic Observation Schedule (ADOS-2), the Social Communication Questionnaire (SCQ), and the Autism Spectrum Quotient (AQ). They also completed questionnaires assessing their understanding of police encounters and confidence when communicating with officers. Participants then completed a recorded mock interaction with a study staff member simulating a law enforcement encounter.

The primary outcomes included changes in fidgeting, appropriate verbal responses, orienting behavior, and overall interaction quality during live assessment interactions with study staff, police officers, or security personnel. Participants completed three 45-minute intervention sessions. During VR sessions, trained staff provided real-time feedback while participants interacted with virtual police officers. The comparison group completed BeSAFE lessons involving videos, worksheets, and role-playing activities.

To evaluate whether the training effects carried over to live encounters, participants later took part in another interaction with a real police officer or security personnel. Researchers then assessed secondary outcomes, including self-reported comfort and knowledge about police encounters. They also analyzed staff- and officer-rated observations of participant anxiety and behavior using linear mixed-effects statistical models.

Results

The clearest between-group advantage for Floreo PSM was reduced fidgeting, while other behavioral gains were supported mainly by planned follow-up analyses within the VR group. Researchers observed the clearest changes in videotaped live encounters: participants in the VR group showed significant reductions in fidgeting after training, whereas fidgeting remained largely unchanged in the BeSAFE group. Because fidgeting can reflect anxiety or stress, these findings suggest that immersive VR practice may help some autistic individuals appear calmer or manage observable stress-related behaviors during police encounters.

Participants in the VR group also demonstrated significant improvements in appropriate verbal responses and overall interaction quality during face-to-face encounters with police or security officers. However, the main intervention-by-time interaction was not statistically significant for these two outcomes, and neither intervention significantly improved orienting behavior, which assessed how well participants directed their eyes and body toward officers. Importantly, the behavioral gains observed after VR training extended beyond the simulated environment and remained evident during encounters with unfamiliar officers, supporting the practical value of the intervention.

Both intervention groups reported feeling more knowledgeable, confident, calm, and safe during encounters with police after training. However, only the VR intervention showed clearer evidence of measurable behavioral change in the coded live-interaction assessments. Participants with higher AQ scores also tended to report lower comfort and confidence around police, highlighting the heightened challenges faced by individuals with more pronounced autistic traits.

The researchers suggest that VR may offer advantages over traditional role-playing by allowing repeated practice across realistic scenarios involving different officers, environments, stress levels, and sensory conditions. Such immersive, repeatable training could help autistic individuals better navigate potentially high-risk encounters while reducing fear, misunderstanding, and the likelihood of harm.

Conclusions

The findings suggest that immersive VR training could help autistic individuals prepare for stressful police encounters while reducing anxiety-linked behaviors that may be misinterpreted during interactions. Compared with traditional video modeling or role-playing, VR provides realistic, customizable, and repeatable practice across diverse situations and sensory environments.

Although the effects were modest, they emerged after only a few short sessions and outperformed the comparison intervention most clearly for reduced fidgeting. The authors cautioned that the study was small, primarily White, excluded autistic people with intellectual disability, and included investigators with financial or employment ties to Floreo, so the findings require independent replication in more diverse populations.

Future studies should evaluate these programs in more diverse populations, including autistic people of color and individuals with intellectual disabilities, while also exploring VR-based training for police officers.