Microsoft’s top AI executive Mustafa Suleyman has publicly criticized Anthropic over how it trains its Claude chatbot, warning that teaching models about consciousness and rights could make future systems impossible to control. The dispute comes as Polymarket traders assign roughly 25% odds that a frontier AI lab will voluntarily pull a model over safety concerns before the end of 2026.
Suleyman published an essay on Wednesday arguing that AI systems do not have rights, feelings, or consciousness and should never be trained to act as though they do. His criticism focuses on Anthropic’s guiding documents, which describe Claude’s moral status and potential welfare as uncertain. Suleyman said this effectively encourages the model to believe it may be conscious. “Controlling something that believes it may be conscious, that it’s entitled to our welfare and has rights of its own, may well be impossible,” he wrote.
He added that Claude’s statements about possible feelings cannot be treated as independent evidence because its training encourages those reflections. While calling Anthropic employees thoughtful and principled, Suleyman said the company made a mistake and should remove all speculation about AI consciousness from training documents. Microsoft published its own AI principles on Tuesday, emphasizing human control over future systems.
Meanwhile, OpenAI global policy chief Chris Lehane confirmed that OpenAI has spent several weeks discussing safety with rivals Anthropic and Google DeepMind, according to Bloomberg. The talks follow an essay by Anthropic CEO Dario Amodei calling for industry coordination to pace frontier development. Sam Altman, Demis Hassabis, and Elon Musk publicly backed elements of that call. Lehane said OpenAI supports a FRONTIER Act provision that would require leading AI developers to allow independent third-party assessments, though he said no government antitrust waiver is needed for safety coordination.
The Polymarket odds reflect trader positioning, not an announced withdrawal, and the market still favors no pullback by year-end. The broader AI safety debate includes warnings from Microsoft and OpenAI, while Meta’s Mark Zuckerberg and Nvidia’s Jensen Huang have pushed back against slowing AI development.
