While not surprising to anyone, I found it interesting that airport security are becoming increasingly reliant on automated AI tools (especially from startups) that utilize travel data such as PNRs. There isn’t really a clear doctrine that explains why someone may be placed on a “high-risk” registrar. Not even the GDPR could completely protect folks living in the EU due to national security exemptions.
Travizory, a Swiss-based startup launched in 2019, sells AI-driven border management software including API-PNR systems that assess travelers and gives them a color-coded risk rating (green, yellow, orange, or red). They also sell biometric exit and entry systems, hoping to combine the two products. “Our goal is to put a facial biometric corridor in [each] country, so people walk through the corridor, and either they are green and they can go to baggage or they are directed to see an immigration, custom, or health officer based on the risk which is identified,” Irminger says.
While Travizory’s system can clear someone to enter a country via machine, those on risk lists have to be further evaluated by a human being. When Travizory’s system flags a person, their name and other information gets sent to the relevant authority along with a tag like “low risk” or “high risk.” Different government agencies in a country—including customs, drug enforcement, immigration, and intelligence services—have different risk classifications, so the system can generate multiple risk lists. Travizory’s AI is trained separately for each country, and the data is stored on AWS, Amazon’s cloud or AWS Outposts in-country.