Imagine the benefits of being able to instantly tell when a person was lying, particularly in a courtroom.
Well, that could soon happen thanks researchers from the University of Maryland (UMD) and a system based on artificial intelligence (AI) and machine learning called the Deception Analysis and Reasoning Engine (DARE).
The way in which DARE works is by looking for micro-expressions in people such as frowning of the eyebrows or protruding of the lips as well as looking at vocal patterns to determine whether a person is lying or not.
However, accurate is not a word the team are using to describe it.
While the AI did perform better than a human at detecting lies in a recent test, there’s still a long way to go before it’s ready to be released commercially.
“An interesting finding was the feature representation which we used for our vision module,” said Bharat Singh, researcher at UMD. “A remarkable observation was that the visual AI system was significantly better than common people at predicting deception.”
There are of course other lie detection technologies in use today which do produce quite reliable results. However, none of them are suitable for a courtroom.
Truth serums are also available but are illegal and polygraphs aren’t usually taken seriously in court. DARE could be the one lie detecting technology to get in and really make a difference.
The team say the technology is not limited to the courtroom and that intelligence agencies could also make use of the AI.
“The goal of this project is not to just focus on courtroom videos but predict deception in an overt setting,” said Singh. “We are performing controlled experiments in social games, such as Mafia, where it is easier to collect more data and evaluate algorithms extensively.
We expect that algorithms developed in these controlled settings could generalize to other scenarios, also.”
However, not everyone is onboard with the AI system yet. “If this is going to be used for deciding [….] the fate of humans, it should be considered within its limitations and in context, to help a human – the judge – to make a decision,” said Raja Chatilla, executive committee chair for the Global Initiative for Ethical Considerations in Artificial Intelligence and Autonomous Systems at the Institute of Electrical and Electronics Engineers (IEEE).
Singh suggests we are just three or fours years away from having AI that can detect lies with a 100% accuracy by reading the emotions that lie behind human expressions.