Airport Facial Recognition: What You Need to Know
TSA has deployed facial recognition at 84 US airports — but most travelers don't know they can refuse it. Here's how the program works, what happens to your data, and how to opt out.
- What is airport facial recognition?
- Is TSA facial recognition mandatory?
- How to opt out at the checkpoint
- What happens to your biometric data
- Accuracy and racial bias concerns
- The oversight gap: what the GAO found
- Congressional pushback
- How other countries handle airport biometrics
What is airport facial recognition?
TSA has deployed facial recognition technology at 84 US airports as of early 2026, as part of a program called CAT-2 (Credential Authentication Technology, version 2). When you reach a TSA checkpoint at one of these airports, a camera captures your face and compares it against the photo on your government-issued ID — your driver's license or passport — pulled from a federal database in real time.
The goal, according to TSA, is faster and more accurate identity verification. If the system confirms a match, you move through without a TSA officer manually inspecting your ID. If it does not match, or if you opt out, an officer performs a manual check.
TSA began rolling out the technology in 2022 and accelerated expansion through 2024 and 2025. For more on how facial recognition works as a technology, see our facial recognition explainer.
Is TSA facial recognition mandatory?
No — TSA facial recognition is not mandatory, and you have the right to refuse it. TSA's own official policy states that travelers may opt out of the facial recognition scan. You will not be denied boarding, flagged for additional screening, or placed on any list for refusing.
The problem is that the opt-out is largely invisible. There is no prominent signage at most checkpoints explaining that refusal is an option. The camera activates as a standard part of the process. Many travelers assume the scan is required.
The Electronic Frontier Foundation (EFF) has documented that TSA checkpoints are structured in a way that makes opt-out the exception rather than the norm, even though TSA's own policy treats it as a routine accommodation.
How to opt out at the checkpoint
You can refuse TSA facial recognition by saying four words to the officer: "I opt out of facial recognition." The officer is required under TSA policy to allow a manual ID check instead. You will not face any penalty.
The EFF recommends being even more specific. Their guidance: tell the officer "I decline the photograph. I want a manual ID check." Being direct reduces any ambiguity about what you are refusing.
- Carry a physical government-issued ID — you need it for the manual check.
- Speak up before the officer directs you to face the camera. Once the scan has already occurred, opting out after the fact has no effect.
- If the officer pushes back or says it is required, ask to speak with a supervisor. TSA policy is clear that refusal must be accommodated.
- You do not need to explain why you are opting out. "I prefer a manual check" is sufficient.
You can also find additional advocacy resources on our fighting back page.
What happens to your biometric data
TSA says the facial image captured at the checkpoint is deleted within 24 hours — but that policy covers only the new scan, not the underlying photo record it was matched against.
This is the detail most coverage misses. The system matches your live scan against a photograph already held in a federal database — the photo from your Real ID driver's license or passport. That photo was submitted to the Department of Homeland Security or a state DMV when you applied for the document. It is part of a permanent federal record. It is not subject to the 24-hour deletion policy.
The "we delete your data" reassurance only applies to the airport-specific capture. Your biometric record — the one the system actually uses to verify your identity — remains in federal systems indefinitely. The American Civil Liberties Union (ACLU) has noted there is no independent audit mechanism to verify even TSA's deletion claims for the airport-captured scans.
Accuracy and racial bias concerns
Facial recognition systems do not perform equally across all demographic groups — and TSA deployed its specific system without completing demographic accuracy testing on its vendor's technology before rollout.
This finding comes from a US Government Accountability Office (GAO) report published in August 2023. The GAO found that TSA had not tested whether its CAT-2 vendor's algorithm performed equitably across racial and ethnic groups before expanding the program nationwide. NIST has published extensive research showing that facial recognition error rates are meaningfully higher for darker-skinned individuals, women, and older adults.
TSA acknowledged the GAO's finding and committed to conducting demographic accuracy testing. As of 2026, no results have been published publicly.
The oversight gap: what the GAO found
The August 2023 GAO report found that TSA expanded its facial recognition program without completing a legally required privacy impact assessment (PIA).
Federal agencies are required under the E-Government Act of 2002 to conduct PIAs before collecting new types of personal information at scale. TSA began expanding CAT-2 to additional airports without completing this assessment. TSA responded by saying it planned to complete the expanded PIA. As of 2025, that updated assessment remains pending.
For a deeper look at AI oversight gaps, see our AI regulation guide.
Congressional pushback
Several members of Congress have formally opposed TSA's facial recognition expansion, though no legislation has passed as of 2026.
Senator Jeff Merkley (D-OR) introduced the Traveler Privacy Protection Act in 2023, which would ban TSA from using facial recognition at airport checkpoints. It passed out of committee but has not received a full Senate vote.
In 2024, Senator Ed Markey (D-MA) and Representative Pramila Jayapal (D-WA) sent a formal letter to TSA demanding a halt to expansion until the GAO's concerns were addressed. TSA did not halt the expansion. Stay current on this and similar developments at our daily AI news briefing.
How other countries handle airport biometrics
US airport facial recognition is part of a broader global trend, but the rules and opt-out procedures vary significantly by country.
- United Kingdom: Heathrow, Gatwick, and other major UK airports use biometric e-gates for passport holders. Travelers can request staffed lanes as an alternative.
- European Union: EU airports are implementing biometrics under the ETIAS framework. GDPR provides data rights, though specific protections at airport checkpoints vary by member state.
- Australia: Australia's SmartGate system uses facial recognition for passport holders at international arrivals. It is designed as optional for most travelers, with staffed counters available.
The common thread: the opt-out exists on paper, but the default experience is the biometric scan. Travelers who do not proactively ask for an alternative will be processed biometrically.
Frequently asked questions
▸ Is TSA facial recognition mandatory?
▸ How many airports use facial recognition?
▸ Does TSA delete your facial recognition data?
▸ Is airport facial recognition accurate for all travelers?
▸ What law allows TSA to collect biometric data at airports?
▸ Can TSA deny you boarding if you refuse facial recognition?
Latest related briefings
AI Layoffs Reversed: Companies Rehire Workers
Workers regain jobs as companies rehire after AI layoffs, raising questions about AI's role in employment and job security.
Read analysis CIVIL RIGHTSAI Bias in Hiring: Workday Lawsuit's Impact on Job Seekers
Job seekers face AI bias as Workday lawsuit reveals potential unfairness in hiring. Understand how this affects your employment.
Read analysis JOBS LABORRAISE US Secures $500M for AI Job Training Programs
RAISE US secures $500M for AI job training, aiming to ease job loss fears. Will these programs effectively secure future employment?
Read analysis