Expert rips ‘irresponsible’ AI study over blackmail scenarios

Anthropic's study tested AI models in simulated environments designed to push edge-case decision-making...

Redirecting to full article...