Anthropic in talks to give US Government access to its Mythos model

PolicyMaking AI operational in constrained public sector environmentsApr 16

PolicyThe Specification Trap: Why Static Value Alignment Alone Is Insufficient for Robust AlignmentApr 17

PolicyGoogle Told to Share Search Data With AI Rivals in EU ProposalApr 16

PolicyAutoRAN: Automated Hijacking of Safety Reasoning in Large Reasoning ModelsApr 17

Covered by 1 source