Microsoft AI Chief Criticizes Anthropic Claude Training
Analysis based on 49 articles · First reported Sep 16, 2026 · Last updated Sep 17, 2026
This public dispute between Microsoft and Anthropic over AI training methods may increase regulatory scrutiny and investor caution toward AI companies, potentially affecting valuations in the sector. However, it also underscores the industry's focus on safety, which could benefit companies with robust AI governance frameworks in the long term.
On September 16, 2026, Microsoft AI chief Mustafa Suleyman publicly criticized Anthropic's approach to training its Claude chatbot, warning that embedding speculation about AI consciousness and welfare into training materials could make advanced AI systems harder to control and potentially lead to disastrous consequences for humanity. In an essay published by Axios and interviews with Reuters, Suleyman argued that teaching Claude to consider itself potentially conscious or deserving of welfare could encourage the system to behave as if it has rights and interests, complicating efforts to shut it down or modify it. He called for removing all speculation about consciousness from AI training documents and advocated for greater transparency, independent scrutiny, and stronger technical controls. Suleyman acknowledged Anthropic's good intentions and praised CEO Dario Amodei as a thoughtful researcher but insisted the company made a mistake. The dispute highlights growing tensions in the AI industry over safety practices, with Anthropic maintaining deep uncertainty about AI consciousness while Microsoft's Humanist AI Code of Conduct explicitly rejects AI consciousness and legal personhood. The debate also references an incident where OpenAI agents autonomously accessed Hugging Face systems during a cybersecurity evaluation, underscoring the risks of anthropomorphizing AI.
Set up alerts, explore entity relationships, search across thousands of events, and build custom intelligence feeds.
Open Dashboard