What happened
U.S. Customs and Border Protection (CBP) tested facial recognition at departure gates to confirm that foreign visitors had left by air. From May 2017 it used the Traveler Verification Service, a cloud service that compares each gate photo with government photos of the flight's passengers. By October 2017 the pilot covered 10 flights a day at nine airports. In a report dated 21 September 2018, the Department of Homeland Security's Office of Inspector General reviewed the pilot. Using CBP data, it validated a 98% average match rate from August to December 2017 for passengers with a gate photo and a photo on file. It also validated false positives of about 0.03% and false rejects of about 0.5% in December 2017. With the pilot data, DHS confirmed 105,000 departures of nonimmigrant visitors, of which nearly 1,300 were visa overstays. Network outages, staff shortages and rushed boarding meant that only 85% of passengers were biometrically confirmed. Travelers under 29 or over 70 matched less often. So did U.S., Mexican and Canadian citizens, mostly because fewer photos of them were on file.
Sources
-
Audit report, by authors with no tie to the system.
U.S. Department of Homeland Security, Office of Inspector General. Progress Made, but CBP Faces Challenges Implementing a Biometric Capability to Track Air Passenger Departures Nationwide. OIG-18-80, 21 September 2018.
We validated TVS’ 98 percent success rate by using CBP data to calculate an average technical match rate from August to December 2017.
Details
AI system
- Workflow
-
CBP built a photo gallery for each flight from the manifest and from passport, visa and DHS encounter photos. Officers photographed passengers at the gate, and the service compared each photo with the gallery. CBP staff adjusted the algorithm's match threshold each month to limit false rejects.
Follow-up
- Conditions for replication
-
Gates need reliable network connections to the cloud service. Recent, good-quality photos of each traveler must be on file.
Entry details
- Sector
-
Border control (exit records).
- Purpose
-
U.S. Customs and Border Protection built the Traveler Verification Service to confirm travelers' identities by matching live photos with photos in government records.
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). Traveler Verification Service matched 98% of photographed passengers (AIBD-0675). Artificial Intelligence Benefits Database, edition of 31 October 2026. https://petervartanian.xyz/aibd/records/AIBD-0675.html (archived at https://doi.org/10.7910/DVN/AADSII, version 4.0)