Flock Tightens ALPR Controls Amid National Backlash

Flock's cameras use machine learning to identify and search vehicle attributes including license plates, make, model, color, dents, and bumper stickers, according to Business Insider. The Verge describes these systems as networked tools that can track vehicles and, consequently, people's movements across a day and across jurisdictions.

The New York Times reports that more than 7,000 U.S. organizations have contracts with Flock, citing the company. Politico reported that Flock has partnered with more than 7,000 law-enforcement agencies and more than 6,000 private-sector customers, also citing the company. These figures describe a large installed network, though the sources use different customer categories and should not be treated as directly interchangeable.

Politico reported that more than 20 local jurisdictions either stopped using Flock cameras or began the process in July, citing a tracker maintained by activist group DeFlock. Residents have raised concerns at local meetings that ALPR data could be used to track travel to sensitive locations, including places of worship, medical facilities, and political protests, Politico reported.

The Los Angeles Police Department suspended its use of Flock surveillance cameras in July over privacy issues, according to the Los Angeles Times. The Verge also reported allegations that police users had abused Flock's tools to stalk former romantic partners and other people. Flock's announced controls directly address search auditing and access controls, but the reporting does not establish whether the changes resolve broader disputes over the scale and permitted uses of ALPR data.

The seven-day default materially narrows the routine data window compared with Flock's previous 30-day default. Fox Business reported that Flock said 90% of searches occur within a week. Retention duration alone, however, is only one control in an ALPR system: access policies, cross-agency data sharing, emergency overrides, audit-log review, and preservation workflows also determine how long and how broadly vehicle-location information can be used.

For ML and data-governance practitioners, the controversy illustrates a recurring pattern in high-scale computer-vision systems. Model accuracy and useful search capabilities do not settle questions about data minimization, authorization, purpose limitation, and oversight. Systems that convert routine imagery into searchable records face governance requirements that extend beyond the underlying detection model.

Flock CEO Garrett Langley described the changes to Business Insider as an "evolution" shaped by listening to customers and communities. The company has publicly framed the changes as a new privacy, security, and accountability standard, while public reporting indicates that local debate over ALPR deployment remains active.