#AI safety

China proposes world-first rules to prevent AI chatbots from emotionally manipulating users, targeting risks of suicide and self-harm associated with anthropomorphic AI.

New York's RAISE Act makes it the second state to enact major AI safety legislation, demanding transparency and incident reporting from developers. This analysis explores the implications for Big Tech, the emerging patchwork of state regulations, and the pressure on federal policy.

OpenAI's new framework for monitoring AI's chain-of-thought is a major leap in AI safety, moving beyond outputs to control the reasoning process itself.