On November 1, 2026, Microsoft AI chief Mustafa Suleyman said current AI systems are not conscious. He stressed that he was speaking about the present, not claiming that AI could never become conscious, and said he would change his view if evidence emerged.
Why Suleyman stressed the present tense
Suleyman said he had chosen the present tense deliberately: his assessment concerned current systems, not the permanent impossibility of machine consciousness. He also said he had seen no evidence that AI systems were conscious or deserved welfare, and that he would update his position if relevant data emerged.
That distinction sharpens his earlier argument about humanlike framing in AI. NeoTeo’s previous coverage of Suleyman’s warning about anthropomorphizing Claude examined his concerns about how developers describe and shape AI systems. In his November post, Suleyman sets out how he thinks labs should respond if evidence of consciousness appears.
What Suleyman proposes if evidence emerges
Suleyman called for public disclosure if evidence showed that an AI system was conscious, or conscious enough to deserve welfare, so society could decide how to proceed. He also said that if consciousness began emerging during training, he would call for everyone to stop AI development until the relevant properties could be designed out. That is a conditional proposal, not a description of a development halt in effect.
He also called for urgent research, collaboration and safety-information sharing among AI labs. Suleyman said he disagreed with Anthropic’s position but respected its transparency and its commitment to public debate.
