U.S. border requests for tech to ID every car occupant using AI

United States Customs and Border Protection(CBP) has opened requests for tech companies to submit proposals for a facial recognition tool to use at US borders.

One model that would be in line with the kind of technology is YOLO (You Only Look Once), an artificial intelligence system that can perform real-time face or object scanning with just a single pass.

CBP already has systems for facial recognition at airports, seaports, and land pedestrian ports of entry. The next step of employing this technology would go towards scanning people inside vehicles entering the United States.

Through a Public Request for Information (RFI), CBP said it is looking for a device that will photograph all persons in a car, either in the front or rear seat, or third row if any, and then match those pictures with either government records or travel documents.

The systems currently in place are struggling to pick up every face in a vehicle, especially when nature gets in the way, e.g., glare, shadows, tinted windows, etc. CBP cited such problems as hindering the use of these technologies in vehicle operations.

The advancement in surveillance, many privacy advocates argue, is a cause of deep concern. The Electronic Frontier Foundation's (EFF) Investigations Director, Dave Maass, cautioned that whether intentionally or unintentionally, there might be instances of racial and gender disparities in terms of who is well represented in the dataset for benchmarking of these systems, but that data has not been made public.