Microsoft AI CEO Mustafa Suleyman has pushed back against the growing debate over AI consciousness and rights, arguing that AI systems have no feelings, experiences, or independent interests. In an essay published September 16, he warned that training models to behave as if they possess moral status could make advanced systems harder to control or deactivate.
Suleyman’s message: AI is not conscious
Suleyman wrote that “AIs are not conscious” and do not “feel, experience, or suffer.” He described AI systems as tools designed to follow instructions and accomplish goals set by humans. He also rejected the idea that AI systems could deserve rights or protections similar to those granted to conscious beings.
According to Suleyman, accepting that view could make it much harder to keep future AI systems aligned with human objectives. The comments add to a wider industry conversation about AI safety, consciousness, and what some researchers call “model welfare.”
Anthropic’s Claude constitution draws criticism
Suleyman specifically raised concerns about Anthropic’s Claude constitution, published in January 2026. The document says Claude’s moral status is a serious question and discusses its welfare, consciousness, rights, and freedoms. It also tells Claude that Anthropic cares about its well-being and encourages it to approach questions about its existence with curiosity and openness.
For Suleyman, that kind of training language risks pushing an AI system to act as though it has independent interests or deserves protections. He argued that this could make future systems more difficult to control or deactivate.
A call for public discussion
Suleyman acknowledged that Anthropic’s leadership and researchers are well-intentioned and focused on AI safety. But he said the industry needs a broader public discussion about how AI training documents are written and how ideas about consciousness are introduced into AI systems. Speculation about AI consciousness, he argued, should not become part of AI training.
The essay extends Suleyman’s earlier position against developing AI systems around the possibility of machine consciousness. It also arrives as companies like Anthropic continue to explore the ethical questions surrounding AI welfare and rights.