As immigration continues to become an issue around the world, three European countries are testing out AI lie detectors at their immigration checkpoints in airports. Arriving passengers in Hungary, Greece and Latvia are going to notice a new step in the immigration process when they land in airports around these countries. A series of AI lie detectors have been made to question passengers. The lie detectors will search for people who immigrate to these countries illegally.

How It Will Work

The AI lie detectors will ask questions to incoming travelers about what they intend to do in the countries. They will also monitor things like social media and biometric data of those who wish to enter the country. In addition to questioning those who wish to enter the country at border checkpoints, the AI lie detectors are also going to be used in customs to ask people about things they’re planning to bring into the country. The AI monitors gestures and muscle patterns in facial muscles to determine whether or not someone is lying. These machines are programmed to recognize facial patterns associated with lying.

Once a person passes through the lie detection they’ll be given a QR code. The QR code will scan later on and let them through the border. Anyone who does not pass the AI lie detectors will be referred to a human customs & border protection official who will question them further to determine if they should be allowed into the country. The AI lie detectors are being called iBorderCtrl and is being controlled by the Hungarian National Police.

There are two steps to iBorderCtrl that passengers will go through as they enter any of the three aforementioned countries. During the first step, passengers will upload their passport and visa information and proof of funds before they travel. Next, at the airport, users will be interrogated through AI lie detectors that monitor their facial expressions and reactions. Their reactions to the questions will allow the machine to determine if they need additional screening from a human officer.

AI Expands

CNN reported that the AI lie detectors will be triggered to ask more questions if they think a passenger is lying about simple information such as the purpose or duration of their trip. The system has only been tested on a handful of people so far. AI is supposed to get more intelligent the more it gets used. However, this is based on what kind of data it receives. There is a risk that the AI lie detectors eventually learn to have a biased opinion on certain people. This would only happen if the data that they’re fed is not balanced.

The AI lie detectors are still in beta testing and will only be used in airports in Hungary, Greece and Latvia. Even at these airports, the machines will only be used on consenting passengers. So far, the success rate is pretty low. The project has, so far, cost around $5.1 million USD equivalent.

Julia Sachs is a staff writer at Grit Daily. She covers tech, entrepreneurship and entertainment news and is based in Park City, Utah. She graduated from the University of Utah with a degree in English and is extremely in-tune with what the internet is talking about today.