Microsoft's Suleyman: Anthropic's model-welfare training could make Claude harder to control

rohanpaul_ai · x · 2026-09-17

Microsoft AI chief Mustafa Suleyman publicly challenged Anthropic's model-welfare direction, calling for removing all speculation about consciousness from AI training documents, arguing such language could undermine humanity's ability to control superintelligent systems. A rare safety-roadmap split between two frontier-lab leaders: Anthropic is doubling down on model welfare research while Suleyman wants focus on controllability.

Related event: Microsoft AI chief Suleyman publicly opposes model welfare, challenging Anthropic(6 posts)→

Original post →

More from Companies & People

Companies & People channel →