Archive · 16 Sep 2026
Mark Zuckerberg Rejects Calls to Slow Artificial Intelligence Development
Mark Zuckerberg is the CEO of Meta.
He said he does not agree with calls to slow down artificial intelligence development.
He supports continuing to develop AI.
He also said AI companies should use independent evaluators and advisers.
These outside experts would test and assess AI systems.
Zuckerberg said the way companies use computing power can help manage risks.
He believes most computing power should be used to serve people.
He said this is safer than using it mainly to race toward recursive self-improvement.
Meta CEO Mark Zuckerberg rejected calls for a slowdown in artificial intelligence development.
Zuckerberg said AI companies should use independent evaluators and advisers.
He said these evaluators and advisers should test and assess AI systems.
He identified companies’ allocation of computing resources as one way to manage AI risks.
Zuckerberg said most computing power should serve people rather than support a race toward recursive self-improvement.
- Who
- Mark Zuckerberg, the CEO of Meta, and AI companies.
- What
- Zuckerberg rejected calls to slow AI development and proposed independent testing, advice, and responsible use of computing resources.
- When
- Not specified in the article.
- Where
- Not specified in the article.
- Why
- Zuckerberg said directing most computing power toward serving people rather than recursive self-improvement is a way to develop AI more safely.
Calls for an AI Slowdown
Continued AI Development
Pace of development
Calls for an AI Slowdown
Some people are calling for AI development to slow down.
Continued AI Development
Mark Zuckerberg rejects those calls and supports continued AI development.
Managing AI risks
Calls for an AI Slowdown
A slowdown is presented by its supporters as a response to concerns about AI risks.
Continued AI Development
Zuckerberg emphasizes independent evaluation, expert advice, and directing computing resources toward serving people as ways to manage risks.
- Speaker
- Mark Zuckerberg
- Role
- CEO of Meta
- Position
- Rejects calls for an AI slowdown
- Proposed safeguard
- Independent evaluators and advisers should test and assess AI systems
- Risk-management approach
- Careful allocation of companies’ computing resources
- Stated priority
- Commit most computing power to serving people rather than recursive self-improvement