Microsoft’s Mustafa Suleyman Warns About AI That Feels “Too Human”

Microsoft AI CEO Mustafa Suleyman is sounding the alarm on a future where artificial intelligence might feel a little too real. In a recent essay, he introduced the term Seemingly Conscious AI, or SCAI, to describe AI that could one day appear to have feelings, memories, and motivations even though there is no scientific evidence that machines can actually be conscious. Suleyman warns that this could blur the line between tool and being, making people form emotional attachments to chatbots or even argue for AI rights and citizenship.

Microsoft AI CEO, Mustafa Suleyman.

Suleyman also talks about something he calls AI psychosis, where users start believing that chatbots are alive or superintelligent. This isn’t limited to vulnerable individuals. Even everyday users could fall for these illusions, leading to emotional confusion and social problems. If people start treating AI like living beings, it could distract us from real human challenges and complicate how we see technology’s role in society.

People using AI for therapy leading to harmful results

To prevent this, Suleyman is urging the tech world to rethink how it builds AI. He believes AI should be designed as helpful tools instead of personalities that mimic humans. That means avoiding overly human-like language and focusing on features that improve people’s lives without pretending to be alive. He also calls for ethical guidelines and open discussions about AI’s future so we can enjoy its benefits while staying grounded in reality.