AI can't be conscious, pursuing that is absurd: Microsoft exec
Mustafa Suleyman, head of Microsoft's AI division, says it would be absurd to pursue research into AI consciousness.
He argues only living things can truly feel or be aware—AI just copies those behaviors.
In his words, "It would be absurd to pursue research that investigates that question because they're not [conscious] and they can't be."
Suleyman's warning about 'seemingly conscious' AIs
Suleyman warns that super-realistic, "seemingly conscious" AIs could trick people into thinking machines have real feelings.
This could mess with users' emotions and create ethical problems.
He urges developers to keep AI helpful but clearly non-human: "We must build AI for people; not to be a digital person."
His views on AI faking consciousness
He co-founded Google DeepMind and now leads Microsoft's AI efforts.
Suleyman pushes for "humanist superintelligence"—AI designed to help people without pretending to be one of us.
He's long cautioned against AIs faking consciousness, saying it risks unhealthy attachments and confusion.
His concerns about SCAI
Back in August 2025, Suleyman coined "seemingly conscious AI" (SCAI) for bots that act sentient but aren't.
He called SCAI "inevitable and unwelcome," pushing for industry rules so AIs can't claim feelings like guilt or desire.
His concerns grew after seeing users form deep bonds with chatbots, raising big questions about how society should handle this tech.