Flock says its new tool will help identify police abuse, but hasn't explained how it works
Flock is reducing default data retention from 30 days to 7 days and introducing "Evidence Mode" for extended retention in exceptional cases requiring a specific case number The company is mandating all customers enable "Audit Assistance" by end of year, a tool that detects and flags abnormal search patterns for administrator review and can auto-lock flagged users Flock confirmed Audit Assistance is not AI or machine learning-based but rather a rule-based data tool that flags atypical patterns su
Analysis
TL;DR
- Flock is reducing default data retention from 30 days to 7 days and introducing "Evidence Mode" for extended retention in exceptional cases requiring a specific case number
- The company is mandating all customers enable "Audit Assistance" by end of year, a tool that detects and flags abnormal search patterns for administrator review and can auto-lock flagged users
- Flock confirmed Audit Assistance is not AI or machine learning-based but rather a rule-based data tool that flags atypical patterns such as searching the same plate with multiple case codes
- Privacy advocates and civil liberties organizations remain skeptical, citing lack of transparency about how the algorithm works, absence of published effectiveness statistics, and no third-party audits
- Critics argue that without legal constraints and warrants, auditing tools alone are insufficient and law enforcement can find workarounds regardless of how effective the tool claims to be
Why It Matters
This development highlights the growing tension between surveillance technology companies and civil liberties advocates as misuse of ALPR systems by law enforcement becomes increasingly documented and public. For AI and surveillance practitioners, it underscores the importance of building transparent, auditable systems rather than relying on opaque proprietary tools that may serve more as public relations measures than genuine safeguards. The case also illustrates how companies in the surveillance space are responding to reputational risk by introducing accountability features, even as experts question whether those features address root causes.
Technical Details
- Audit Assistance is a rule-based data tool, not AI/ML-based, that flags atypical search patterns such as a user searching the same license plate with multiple different case codes or repetitively searching the same plate for over 30 days
- The tool automatically locks out flagged users until an administrator intervenes, and provides a documented workflow for reviewing atypical activity
- Over one-third of Flock's customers had voluntarily enabled Audit Assistance since its April launch, and it is now being made mandatory for all customers
- "Evidence Mode" allows extended data retention beyond 7 days in exceptional cases but requires a specific case number to activate
- Customers can now set data-sharing limits with other customers based on the type of offense under investigation
- No published statistics on false positive/negative rates, no disclosed training data, and no third-party audit results have been made available
Industry Insight
- Surveillance technology vendors face increasing pressure to demonstrate accountability and transparency; self-regulatory tools without independent verification risk being dismissed as "window dressing" and may accelerate demands for legislative oversight
- The lack of transparency around how Audit Assistance works — including whether it uses AI, what patterns it detects, and its effectiveness metrics — represents a significant credibility gap that could affect customer trust and regulatory scrutiny across the surveillance industry
- The recurring pattern of law enforcement abuse of ALPR systems suggests that technical controls alone are insufficient; companies in this space should anticipate and prepare for stricter regulatory requirements, including warrant mandates and independent auditing, rather than relying on proprietary auditing features as a substitute for legal safeguards
Disclaimer: The above content is generated by AI and is for reference only.