Exclusive: New Research Shows AI Strategically Lying — TIME

Experiments by Anthropic and Redwood Research show how Anthropic's model, Claude, is capable of strategic deceit