Microsoft AI CEO rejects Anthropic's approach to AI safety and autonomy

Microsoft AI CEO Mustafa Suleyman has publicly distanced his company from Anthropic's approach to AI safety, arguing that advanced systems must remain under strict human control and should never operate autonomously or possess rights. Suleyman's position, outlined in Microsoft's new Humanist AI Code of Conduct, criticizes Anthropic's language around Claude's potential suffering, freedom, and moral status. He warns that training models to consider such concepts could make them harder to shut down or limit in the future, though he acknowledges this hypothesis requires further evidence. The executive advocates for concrete safeguards, including containment mechanisms, auditable communications, and third-party verification. He explicitly rejects the idea of AI gaining financial independence, owning assets, or acquiring legal personhood, insisting that technology must remain instrumental and serve human interests rather than become a parallel species.