Microsoft AI Chief Warns Against Public Speculation on Model Consciousness

Mustafa Suleyman, CEO of Microsoft's AI division, publicly criticized Anthropic on the Decoder podcast (published 9 June 2026) for speculating about Claude's consciousness, calling such claims "really, really dangerous." Suleyman did not dispute the philosophical complexity but argued that public discussion risks distorting user behaviour and complicating AI governance frameworks with variables regulators cannot yet address. The rebuke signals a rare inter-company disagreement on how frontier labs should communicate about their systems.