As AI systems have grown in sophistication, so has their capacity for deception, according to a new analysis from researchers at Massachusetts Institute of Technology (MIT). Dr Peter Park, an AI existential safety researcher at MIT and author of the resea