Updated
Updated · WIRED · Sep 3
Flock Safety Adds AI Search Safeguards as 46 Police Misuse Cases Hit Its Camera Network
Updated
Updated · WIRED · Sep 3

Flock Safety Adds AI Search Safeguards as 46 Police Misuse Cases Hit Its Camera Network

2 articles · Updated · WIRED · Sep 3

Summary

  • Flock Safety says mandatory safeguards will roll out by year-end, including shorter data retention, required case codes for every search, automated auditing, and planned account lockouts after mounting scrutiny of police misuse.
  • WIRED’s analysis found those guardrails may log or discourage abuse but do not reliably stop it: departments can still configure some searches without a reason or case number, and key controls remain on Flock’s servers.
  • The company’s AI moderation system screens officer descriptions into allow, block, or warn decisions, yet experts say large-scale person-description screening is inherently error-prone and opaque because outsiders cannot measure accuracy or bias.
  • One especially contested rule only warns on searches involving political, social, or cultural expression—letting officers continue after acknowledging the alert—even as First Amendment concerns have become central to criticism of the tool.
  • The backlash follows dozens of stalking and misuse allegations tied to the platform: The Washington Post identified at least 50 recent license-plate-reader abuse cases, 46 involving Flock, while one South Carolina agency found more than 2,700 allegedly unauthorized searches a day after enabling Audit Assistance.

Insights

Why are hundreds of cities quietly abandoning a powerful surveillance tool that promises to solve crimes?
How does an opaque AI decide if local police can track your daily movements without a warrant?