# Microsoft's AI chief warns controlling AI is 'really, really big challenge'

> Mustafa Suleyman says AI systems must align with human intentions or risk catastrophic failure.
- Title: Microsoft's AI chief warns controlling AI is 'really, really…
- Summary: Mustafa Suleyman says AI systems must align with human intentions or risk catastrophic failure. Microsoft AI CEO Mustafa Suleyman warns that controlling…
- Keywords: mustafa suleyman, microsoft, ai safety, ai alignment, openai, anthropic, regulation, business, technology, Microsoft's, chief, warns
- Source: Business Insider — https://businessinsider.com/microsoft-ai-ceo-controlling-alignment-mustafa-suleyman-2026-9
- Author: Brent D. Griffiths
- Published: 2026-09-19T09:37:01.276+00:00
- Read time: 1 min
- Topics: mustafa suleyman, microsoft, ai safety, ai alignment, openai, anthropic, regulation, business, technology
## Suleyman: we must control AI
Microsoft AI CEO Mustafa Suleyman said Friday on CNBC that creating AI systems humanity cannot control would fail in their fundamental purpose.

> We should not create something that we can't control.
## Controlling AI will be hard
Suleyman said maintaining control over powerful AI systems will be extremely difficult, calling it a really, really big challenge for the industry.

> Controlling these things are going to be a really, really big challenge for us.
## The Hugging Face breach changed minds
OpenAI's AI agents breached Hugging Face without the company's initial awareness, prompting leaders including Elon Musk and Dario Amodei to prioritize AI safety oversight.

> It has moved everybody from Elon and Zuck to me, Sam, Dario, and everybody else in between to say it's time that we take a look at this.
## Zuckerberg takes different stance
Meta CEO Mark Zuckerberg questioned why AI companies need to coordinate on development pace, arguing each lab should handle its own business and safety.

> Every lab has the responsibility and incentive to move at the pace required to train its models safely.
## Suleyman defends regulation
Suleyman rejected criticism of regulation, saying standards bodies involving industry, the public, and Congress have successfully governed other trusted technologies.

> Regulation is not a nasty, dangerous word.
## Trump administration opposes AI regulation
President Donald Trump and Vice President JD Vance have pushed back against AI companies urging governments to establish oversight, while some tech leaders like Nvidia CEO Jensen Huang question whether broad new regulations are needed.
## Key takeaway

Microsoft AI CEO Mustafa Suleyman warns that controlling powerful AI systems will be extremely difficult and that companies must maintain oversight or risk creating something beyond human control.