What happened
Austin's City Council let the Austin Police Department (APD) run automated license plate readers in a one-year trial from March 2024. Flock Safety supplied 40 fixed cameras, and Axon cameras on patrol vehicles went live in the third quarter. The cameras checked plates against hotlists, and APD kept the data for seven days. In an audit published in May 2025, the City Auditor checked APD's figures for March to December 2024 against Flock data and stop forms. The auditor found that the program led to 165 arrests, about 133 prosecutions and the recovery of a missing or endangered person. Across 13,122 hotlist matches and 286 stops, it found no incorrect or unjustified stop. The prosecution counts came from public information requests and are estimates. Staff did not pull scan data before Flock deleted it after seven days, so the auditor could not verify the Flock scan totals. The auditor also found that the Flock contract may leave motorists' data open to unintended use. The audit makes no comparison with policing without the readers.
Sources
-
Audit report, by authors with no tie to the system.
City of Austin Office of the City Auditor. APD License Plate Reader. Audit Report, May 2025. Finding 1, Resolution item C and Exhibit 3, pp. 6-7. The auditor analyzed Flock ALPR data and APD Risk Management's quarterly audits for March to December 2024.
APD's ALPR program resulted in 165 arrests, approximately 133 prosecutions, and 1 missing/endangered person recovered.
We found that APD's ALPR program used 3 hotlists, found 13,122 matches, and resulted in 0 incorrect or unjustified stops.
Due to a lack of understanding of the deletion period and scheduling failures, all original data was not pulled and stored across all three quarters of the trial period.
Details
AI system
- Workflow
-
Forty fixed Flock cameras from March 2024, and Axon cameras on patrol vehicles from the third quarter, scanned plates and checked them against NCIC, NCMEC and APD hotlists. Officers filed a Hit Stop Form for every stop made on a plate-reader hit, and APD Risk Management audited the program every quarter.
Follow-up
- Conditions for replication
-
Officers may use the data to detain or arrest only for felonies, Class A or Class B misdemeanors and Class C and D hate or sex crimes. Scan data must be pulled before Flock deletes it after seven days, or the program's figures cannot be verified.
Entry details
- Sector
-
Policing.
- Purpose
-
Flock Safety and Axon built automated license plate readers to photograph passing plates and alert police when a plate matches a hotlist.
31 more questions not yet checked
model, release, provider, role, configuration, inputs and outputs, autonomy, human oversight, AI contribution, beneficiaries, reach, distribution and access, duration, every measurement question, every money question, ownership and control, dependencies, cross-border connections, geopolitical implications, replication attempts, availability, institutional decision, grounds, action and result, success criterion, every evidence question, observed or intended, alternative assessments, related entries, other identifiers, category mappings.
Cite this record
Peter H. Vartanian. (2026). Austin's license plate readers led to 165 arrests in a 2024 trial (AIBD-1141). Artificial Intelligence Benefits Database, edition of 1 November 2026. https://petervartanian.xyz/aibd/records/AIBD-1141.html (archived at https://doi.org/10.7910/DVN/AADSII, version 7.0)