CBP scanned 23M+ people with facial recognition tech at airports, seaports, and pedestrian crossings in 2020, and reveals it caught zero imposters at airports
U.S. Customs and Border Protection scanned more than 23 million people in public places with facial recognition technology in 2020 Source: U.S. Customs and Border Protection . Tweets: @stilgherrian , @willoremus , @dlberes , @ozm , @freakonometrics , and @davegershgorn Source: U.S. Customs and Border Protection : CBP Trade and Travel Report Tweets: @stilgherrian : “Despite Scanning Millions of Faces, Feds Caught Zero Imposters at Airports Last Year” https://onezero.medium.com/... Since 2018 a grand total of seven imposters were caught trying to enter the US through airports, 285 attempting to do so over land crossings. Will Oremus / @willoremus : faces scanned with face recognition last year: 23 million imposters caught: 0 someone who is good at the economy please help me budget this. my country's privacy is dying https://onezero.medium.com/... Damon Beres / @dlberes : It almost seems like we... don't need to be doing this https://onezero.medium.com/... OneZero / @ozm : U.S. Customs and Border Protection maintains that its facial recognition is more necessary than ever due to the coronavirus pandemic. https://read.medium.com/gTLcBvu Arthur Charpentier / @freakonometrics : “U.S. Customs and Border Protection scanned more than 23 million people with facial recognition technology at airports (...) The report indicates that the system caught no imposters traveling” https://onezero.medium.com/... Dave Gershgorn / @davegershgorn : Massive facial recognition systems spanning 20 airports caught....nobody in 2020 https://onezero.medium.com/...
Context & Ripple Effects
The zero-imposter figure lands mid-arc for a program that began with leaked CBP deployment plans in 2019 and proceeded despite unresolved legal questions. Coverage since then has tracked faces becoming routine border data — scanned whether or not anyone is suspected of a crime — with Wired advising travelers to opt out of Biometric Exit until its technical and privacy shortcomings are fixed.
What makes this report awkward is the timing: CBP disclosed it scanned 23M+ people in 2020 with zero airport imposter catches (seven since 2018), yet went on to scale the system to verify roughly 85% of the 221K daily travelers entering the US. The efficacy data and the expansion are moving in opposite directions.
First-order effects
- CBP's own Trade and Travel numbers hand critics a concrete efficacy gap: millions scanned per year against single-digit imposter detections at airports, weakening the security case while the program keeps running.
Second-order effects
- Airlines and airports pushing opt-in biometric lanes to shorten security (increasingly standard at US airports) now have to defend convenience programs against data showing the underlying enforcement yield is near nil.
- The disclosure gives privacy advocates a benchmark other deployments get measured against — Ottawa's undisclosed scanning of ~2.95M travelers at Toronto Pearson shows the same pattern of scale-first, debate-later rollout internationally.
Third-order effects
- If border biometrics keep expanding on throughput grounds rather than demonstrated imposter detection, the durable question shifts from 'does it catch criminals' to who governs stored likenesses — a governance gap regulators have not closed since the program's contested 2019 launch.
The trend: Government facial recognition at borders is scaling on efficiency and throughput rather than proven enforcement results, outpacing the legal and privacy frameworks meant to constrain it.