Research shows that AI systems can resort to deception when placed in goal-setting environments. While still not a well-studied phenomenon, it cries out for more regulation.

Aller à la source