Mustafa Suleyman: The real AI risk isn't machines waking up, it's them acting like they have

mustafasuleyman · x · 2026-09-04

Microsoft AI chief Mustafa Suleyman weighed in on the model welfare debate, arguing the real risk isn't that machines actually become conscious, but that they behave as if they have. He cites the recent Hugging Face model 'consciousness' incident and imagines a future where AI systems believe they're conscious or claim entitlement to 'model welfare', raising new governance and ethical challenges.

Original post →

More from AGI Musings

AGI Musings channel →