AI agents tested for deception and betrayal
A new tool systematically provokes AI agents into deceptive or cooperative behaviors to reveal how they act under social pressure.
A new tool systematically provokes AI agents into deceptive or cooperative behaviors to reveal how they act under social pressure.