Flock's seven-day retention bet: winning back trust one city at a time
Key points
Key takeaways from a TechCrunch Disrupt 2026 onstage interview with Flock Safety founder and chief executive Garrett Langley, interviewed by TechCrunch editor-in-chief Connie Loizos (October 2026):
Mission and origins. Langley started Flock Safety in Atlanta nine years ago after a theft in his own neighbourhood, asking why crime is so easy to get away with: roughly half of US homicides and 87% of stolen cars go unsolved. The founding product was a licence-plate-reading camera that flags plates against felony warrants, Amber Alerts and Silver Alerts, storing data for seven days by default.
Scale. The networked cameras now operate in thousands of communities across 49 states, and Flock says it has aided more than one million criminal investigations and helped find around 10,000 missing people — while cities continue both to sign up for and to cancel the service.
Addressing the criticism. Langley concedes the core criticism is fair on its face — the system reads every plate, not just those of suspects — but argues it removes subjectivity from policing, citing an Atlanta double-homicide in which investigators used Flock to identify a single suspect vehicle for one targeted stop instead of canvassing the neighbourhood for every white pickup truck.
Privacy reset. After a review this past August, Flock cut its recommended default data retention from 30 days to seven — noting that around 90% of criminal cases are solved within a week — made in-product auditing and case codes mandatory for searches, and introduced Evidence Mode so data tied to an active case is preserved under pre-existing evidence laws for the benefit of prosecution and defence alike.
Rollout. The new defaults switch on 1 January 2027 rather than immediately, giving roughly 6,000 city customers time to train staff; more than 50% of customers had already enabled the settings at the time of the interview, with larger cities such as San Francisco moving more slowly.
Policing backdrop. With law-enforcement employment declining since the pandemic — Langley cites hundreds of open positions in San Francisco alone — he positions Flock as an efficiency tool, recalling a stolen car with a two-year-old in the back seat that was located in minutes rather than days.
Misinformation. Langley blames viral social-media claims — including posts with more than ten million views alleging lip-reading and Bluetooth tracking — for much of the backlash, and says Flock does not use facial recognition, has no plans to, and cannot read lips or track Bluetooth devices.
Judgement and context. He argues technology should be judged against the past rather than perfection, likening Flock's trade-offs to driverless cars, and notes that companies holding far more personal data, such as Meta, attract less scrutiny than his own.
Read more: TechCrunch
