EU to pilot AI facial analysis to catch dishonest travelers

The European Union will run a pilot of an artificial intelligence border control system for six months at four border crossings between Hungary, Latvia and Greece and non-EU countries, The Verge reports.

The program is the first field test of iBorderCtrl, an EU-funded project to facilitate border crossings with AI. Travelers fill out an online application and upload documentation, such as passport images.

Based on this information, the system’s virtual border guard asks questions, and the traveler’s responses are analyzed for micro-gestures that indicate whether or not the person is answering truthfully. A previous iteration of the technology had a 76 percent success rate in early testing, but a representative of iBorderCtrl told New Scientist that they are confidant it can reach 85 percent.

During the pilot, system judgements will not be used to deny border crossings, according to the report.

The system is developed by European Dynamics, and was one of several projects to receive funding early last year. Similar technology is also being developed to reduce bank loan fraud in China, according to recent reports.

Identity and migration is one of the four strategic challenges identified in the EAB’s new strategic plan, which was launched at the organization’s recent Research Project’s Conference.

Article Topics

artificial intelligence  |  border security  |  EU  |  pilot project