Archive · 16 Sep 2026
Zuckerberg Rejects Calls for Coordinated AI Development Slowdown
Mark Zuckerberg says AI companies should decide for themselves how to develop AI safely.
He does not think companies and countries need to agree on one shared slowdown.
His comments came after Anthropic CEO Dario Amodei proposed coordinated action.
OpenAI CEO Sam Altman and Elon Musk were reported to support that proposal.
Zuckerberg says AI companies have reasons to prevent their systems from causing harm.
He also says companies could delay or stop a release if a model is unsafe.
Meta waited several months before releasing its Muse AI agent to improve its safety and security.
Zuckerberg wants companies to use most of their computing power to help people.
He warned against racing toward AI systems that can improve themselves automatically.
Meta CEO Mark Zuckerberg said each AI company should independently ensure its technology is safe.
His comments followed Anthropic CEO Dario Amodei’s proposal for companies and countries to coordinate a slowdown in advanced AI development.
OpenAI CEO Sam Altman and Elon Musk were reported to support Amodei’s proposal.
Zuckerberg said Meta delayed releasing its Muse AI agent for several months to improve its safety and security.
He urged companies to prioritize serving people instead of racing toward recursive self-improvement in AI systems.
- Who
- Meta CEO Mark Zuckerberg, Anthropic CEO Dario Amodei, OpenAI CEO Sam Altman, and Elon Musk.
- What
- A debate over whether AI companies and countries should coordinate a slowdown in advanced AI development.
- When
- Zuckerberg posted his comments days after Amodei proposed a three-step AI safety plan.
- Where
- Zuckerberg made the comments in a post on X; the report was datelined New York.
- Why
- The debate concerns warnings that advanced AI could threaten humanity or escape human control, and how companies should reduce those risks.
Supporters of Coordinated AI Controls
Zuckerberg’s Company-by-Company Approach
Whether development should slow collectively
Supporters of Coordinated AI Controls
Dario Amodei proposed that AI companies and countries coordinate a slowdown in advanced AI development; Sam Altman and Elon Musk were reported to support the proposal.
Zuckerberg’s Company-by-Company Approach
Mark Zuckerberg said each AI company has the responsibility and ability to move at the pace needed to train and deploy its models safely.
How safety should be enforced
Supporters of Coordinated AI Controls
Supporters of coordination favor joint action to address potential threats from advanced AI before systems could escape human control.
Zuckerberg’s Company-by-Company Approach
Zuckerberg said companies already have incentives and significant liability to prevent their systems from causing harm, and can delay or halt unsafe releases themselves.
What development should prioritize
Supporters of Coordinated AI Controls
A coordinated slowdown would place collective limits on the pace of advanced AI development.
Zuckerberg’s Company-by-Company Approach
Zuckerberg said companies should devote most computing power to serving people instead of racing toward recursive self-improvement.
- Main speaker
- Mark Zuckerberg, CEO of Meta
- Coordinated proposal
- Dario Amodei proposed that AI companies and countries coordinate a slowdown in advanced AI development.
- Reported supporters
- OpenAI CEO Sam Altman and Elon Musk were reported to support Amodei’s proposal.
- Zuckerberg’s position
- Each AI company should take responsibility for developing its technology safely and can delay deployment if necessary.
- Meta safety decision
- Meta delayed releasing its Muse AI agent for several months to improve its safety and security.
- Development priority
- Zuckerberg urged companies to focus computing resources on serving people rather than recursive self-improvement.
