An experiment by researchers at the Georgia Institute of Technology is believed to be the first detailed examination of robot deception.
The team developed computer algorithms that would let a robot ‘decide’ whether it should deceive a human or another robot and gave it strategies to give it the best chance of not being found out.