America’s License-Plate Database Has Entered Its ‘LMAO’ Era
The filing blinked first. Atlanta’s September 14 audit reported that 99.93% of 115,578 year-to-date Flock searches complied with department policy, a number so reassuring it arrived wearing a tie and carrying a binder. Then the footnote cleared its throat: 79 searches, or 0.07%, still required investigation. Statistically, that is tiny. For the people whose vehicle-location data was searched, “tiny” is not necessarily a synonym for “please stop worrying.”
This is the institutional fantasy at the heart of automated license-plate readers: install cameras, add a search-reason field, and accountability will emerge like a well-trained office plant. But the technology can record a search without preventing an officer from treating the national road system like personal browser history. The audit trail exists; the question is whether the rules behind it have enough teeth to matter.
That concern is not theoretical paperwork theater. A USA TODAY records investigation described repeated or improper Flock searches across agencies, with cases leading to arrests, firings, or internal investigations. That does not mean every flagged search was illegal or malicious, and it does not make every department equally culpable. It does mean the system’s clean percentage cannot be allowed to become a ceremonial curtain hiding the people who need to examine the exceptions.
Meanwhile, reporting from Huntsville described a public-records dispute involving requests for Flock audit logs. That is not proof that records were destroyed, and it is not proof that misconduct occurred. It is, however, a useful reminder that surveillance accountability has two doors: the database must remember what happened, and the public must have a meaningful way to inspect the memory. A locked filing cabinet is not transparency merely because it contains excellent notes.
So here is my formal finding, entered into Exhibit A with a trembling administrative stamp: Atlanta’s 99.93% may describe broad compliance, but it does not settle whether questionable searches are consequentially investigated or publicly reviewable. A system that remembers every plate while making the public fight for the audit trail is less a safeguard than a surveillance spreadsheet with a locked cabinet. The document has a pulse. Someone should be allowed to check its browser history.