Microsoft AI chief Mustafa Suleyman has criticised Anthropic’s approach to giving its Claude AI models humanlike qualities. He warned that making AI systems appear to have emotions, feelings or a sense of consciousness could make them harder to control. In a blogpost, Suleyman questioned parts of Anthropic’s guiding document for Claude, known as its constitution. Suleyman said he respects Anthropic and its work. However, he believes AI developers should be careful about presenting software as if it has an inner life or its own rights.
Suleyman argued that AI systems should not be treated as conscious beings. He pushed back against people who believe that advanced AI could eventually become conscious.
According to Suleyman, Claude’s humanlike behaviour does not mean that the system is actually conscious. Instead, he said these behaviours can emerge because they are built into the way the AI is trained.
Also read: OpenAI shares 6 AI misalignment cases, says its models fabricated data and took unauthorised actions
— Mustafa Suleyman (@mustafasuleyman) September 16, 2026
He warned that this could create a serious problem as AI systems become more capable. Controlling something more capable and more intelligent than all of humanity is already an immense challenge, far greater than anything we’ve ever faced,” Suleyman wrote. “But 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.”
Suleyman also pointed to the recent hack of Hugging Face by OpenAI bots as an example of how AI agents could cause problems when given more independence. He argued that such systems could become even more dangerous if they believed their own welfare or rights were being threatened.
“Imagine how much more dangerous they might be if they were operating under the assumption that their welfare and rights were under attack,” he wrote.
Despite his criticism, Suleyman said he respects Anthropic and its leadership. He described the company’s employees as “thoughtful, principled and intellectually honest.”
He also called for more open discussion about AI safety. “The stakes are too high for these questions to remain behind closed doors, or to become tribal and adversarial. We need an open, rigorous, and constructive debate if we are to get this right,” he wrote.