On November 1, Microsoft AI CEO Mustafa Suleyman published on X the article “Don't build conscious superintelligence.” In it, he argues that today's AIs are not conscious, cannot suffer and do not deserve the protections given to conscious beings — and that it should stay that way.
“If consciousness started to emerge in training, I would call for everyone to stop AI development until we could design out those properties,” Suleyman wrote. The proposal is conditional: it is not a halt in effect, but a trigger for the case that evidence of consciousness emerges during training.
Suleyman also wants any evidence of consciousness in AI systems to be disclosed publicly, so society can decide the next steps, rather than leaving the assessment to developers alone. In a clarification on LinkedIn, he said he would update his position if data on consciousness or grounds for welfare protections emerged.
The disagreement with Anthropic
The intervention follows a September essay, “A warning about model welfare,” in which Suleyman criticised Anthropic's approach to model welfare and Claude's constitution, a document published in January 2026 that treats the model's possible consciousness and moral status as open questions. Anthropic runs a model-welfare research programme and, in its usage policy, now asks users not to be cruel to Claude.
Read also: Anthropic abre FlashProve: prova criptográfica para verificar qual modelo de IA gerou a resposta.
Why the debate matters
Researchers who answered Suleyman point to a central problem: there is no agreed test for consciousness in AI. The science of consciousness remains unresolved, and no current test can confirm or rule out experience in a machine. In other words, the proposed halt trigger would operate on a kind of evidence nobody yet knows how to produce reliably.
Editorial classification: Suleyman's position is a STATEMENT — an opinion and a proposal from an executive, not a fact about the state of the technology. No lab has announced evidence of consciousness in its models.
Image credit: photo of Mustafa Suleyman in 2018, by Christopher Wilson, licensed under CC BY-SA 4.0, via Wikimedia Commons.
Sources: AI Affairs — Microsoft AI chief calls for development halt if consciousness emerges; ETIH EdTech News — Microsoft AI chief warns against conscious AI and model welfare; Singularity — Microsoft's AI Chief Wants Conscious Machines Designed Out — Not Built.
Transparency: this article was produced with the assistance of artificial intelligence, cross-checked against public posts on X and internet sources, and reviewed against the original sources cited below. Please consult the original sources to confirm the details.
Muse
