Flock Unveils Tool to Detect Police Misconduct, Yet Details on Functionality Remain Unclear

Flock, a company specializing in surveillance technology, announced new regulations and tools aimed at reducing misuse of its automated license plate recognition systems while ensuring accountability among law enforcement users.

The firm now advises police departments to store data for a duration of seven days, a notable reduction from the former 30-day period. Flock has also introduced a feature called “Evidence Mode,” which enables customers to extend data retention in exceptional circumstances, provided they input a specific case number. Additionally, police departments can customize the extent of data shared with other agencies based on the offenses being investigated.

Among these updates is a tool named “Audit Assistance,” which was rolled out in April. Flock reports that over a third of its clients have started using the new feature, and the company now mandates that all customers activate it by year’s end.

According to Flock, Audit Assistance identifies unusual user activity and flags it for administrative evaluation. If flagged, the system can temporarily restrict access to the user until reviewed by an administrator. The company claims that this measure has already been instrumental in detecting and preventing misuse.

Executives have praised the effectiveness of Audit Assistance, though the specifics of its operation remain somewhat ambiguous.

For example, Ashley Haber, Flock’s head of trust and compliance, stated in a video that the tool can identify “unusual search behaviors from specific users,” aiding customers in monitoring system utilization. Meanwhile, co-founder Paige Todd noted that the feature can detect abnormal patterns, such as a user repeatedly searching the same license plate for over a month.

Major Patrick Krieg of the Dunwoody, Georgia Police Department, which is a client of Flock, referred to the feature as an algorithm that alerts them to potential biases.

Flock portrays Audit Assistance as a tool that can swiftly identify and document unusual activities for immediate review.

However, it remains unclear how this algorithm functions, what data informs it, and whether it employs large language model technology. Questions about the type of patterns it is designed to identify and the criteria for labeling activities as “abnormal” are still unanswered.

Flock clarified that the Audit Assistance tool does not rely on machine learning or artificial intelligence. The company explained that it flags atypical search patterns for further scrutiny, citing an example of a user searching for the same plate under varied case codes, which might suggest potential misuse. However, a flag does not imply confirmed abuse but simply denotes a need for additional investigation.

Despite Flock’s explanations, doubts linger regarding the effectiveness of the tool. Critics, including privacy advocates, have expressed skepticism about Audit Assistance. Chad Marlow from the ACLU noted that there is no substantial evidence supporting the tool’s reliability in curbing abuse. He emphasized the importance of understanding how many officers have misused the system to gauge whether the auditing process is genuinely effective.

Cooper Quintin from the Electronic Frontier Foundation underscored that the tool’s functionality may be irrelevant if law enforcement finds ways to bypass it. He remarked that what truly matters is whether officers face consequences for their actions, suggesting that legal restrictions on technology use might be more effective than reliance on internal auditing.

Rachel Levinson-Waldman from the Brennan Center acknowledged that making Audit Assistance mandatory is a positive step, but called for an independent audit of the tool to clarify its practical impact.

These announcements from Flock follow a history of reported abuses connected to its systems. Recently, three former deputies from a Georgia sheriff’s office were arrested for allegedly using Flock cameras to stalk personal acquaintances, underscoring the need for the newly implemented Audit Assistance tool, which officials claimed was instrumental in uncovering this misconduct.