AI is no longer just an idea. It is an already part of how we live each day. In the background, it silently streamlines and expedites everything from our scheduling to our shopping and television viewing. The fact that individuals are beginning to rely on AI for purposes other than everyday assistance is remarkable. Nowadays, a lot of people use these systems more like trusted friends than like tools, sharing personal information, preferences, and even emotions with them.
Now in a recent blog and follow-up posts on X, Microsoft AI CEO Mustafa Suleyman raised alarms over what he terms “Seemingly Conscious AI” (SCAI). He described these systems as so advanced that they might imitate consciousness to the point where users start treating them as real, sentient entities. “This isn’t about whether AI is truly conscious,” Suleyman explained. “It will seem conscious, and that illusion is what will matter in the near term.”
READ: Meta freezes hiring in AI division after a spending spree: WSJ
Suleyman stressed that the real risk doesn’t come from the technology itself but from the way people might react to it. Once individuals begin treating AI as if it has emotions or moral standing, it could unsettle social norms and institutions. “Many people will start to believe in the illusion of AIs as conscious entities so strongly that they’ll soon advocate for AI rights,” he cautioned.
Suleyman described this AI psychosis as a “real and emerging risk.” On X, Suleyman stressed that there is “zero evidence of AI consciousness today.” He cautioned that the real danger lies in people treating AI as if it were conscious, falling for an illusion that doesn’t exist. Suleyman also attacked businesses that make allegations about the sentience of their systems, stating that neither the AI nor its developers should ever promote such a notion.
READ: GPT-6 preview: OpenAI’s big step toward personalized AI
Suleyman urged the tech sector to establish unambiguous lines that separate AI from humans in order to prevent deceiving users. He maintained that while these institutions play useful, secure, and sustaining functions, they should constantly draw attention to their limitations rather than fostering delusions of individuality. “AIs cannot be people – or moral beings,” he wrote. “We must build AI for people; not to be a digital person.”
For Suleyman, the real promise of artificial intelligence lies in making everyday life easier and more meaningful, not in convincing people that machines are alive. Instead of claiming to be something it isn’t, he thinks AI should be created to encourage human innovation, improve interpersonal connections, and eliminate needless complexity. As he put it, “Sidestepping SCAI is about delivering on that promise, AI that makes lives better, clearer, less cluttered.” His message is clear: the value of AI is in how it serves people, not in how closely it can mimic being one.