What happened
On 12 Sep 2026 Anthropic CEO Dario Amodei published 'We Must Pace the Frontier' urging a speed limit on frontier AI, with OpenAI's Altman, xAI's Musk and DeepMind's Hassabis publicly backing it within hours.
On 12 September 2026, Anthropic CEO Dario Amodei published a 3,800-word essay titled 'We Must Pace the Frontier' on his personal website. His argument was simple: the frontier of AI is moving too fast to properly align and safeguard models before they become self-improving. Within hours, Sam Altman of OpenAI, Elon Musk of xAI and Demis Hassabis of Google DeepMind publicly said they agreed and would pace development too.
Amodei's plan has three steps. Step one is already happening: Anthropic will give independent third-party evaluators employee-level access inside its labs to test models before release, and Altman said OpenAI 'will do the same' with more to share soon. Step two is industry-wide coordination among democratic countries to set common safety standards and a speed limit on unchecked progress. Step three is narrower global coordination, including with authoritarian governments, on shared risks.
The timing matters. The essay landed the same week OpenAI and Anthropic are reportedly circling IPOs, and just weeks after more than 1,000 AI lab employees signed a July letter demanding a slowdown and Anthropic researcher Jacob Coxon's resignation post — 'Neither company is acting responsibly. They are racing straight to self-improving superintelligence and gambling with our lives' — went viral to over 170 million views on X.
Not everyone is applauding. On 14 September, China's Foreign Ministry spokesperson Guo Jiakun called the slowdown talk 'fearmongering,' while President Trump posted that the only guardrails AI needs is a 'STRONG AND SMART (High IQ!) PRESIDENT.' WIRED reported OpenAI has quietly checked with Congress whether a coordinated pause would breach the Sherman Antitrust Act — because rivals jointly restricting output is exactly what antitrust law calls a cartel. Watchdog Sacha Haworth warned, 'We should not be letting the foxes run the henhouse.'
Why the cartel question?
Under US antitrust law, competitors jointly agreeing to limit output or innovation can violate the Sherman Act. That is why OpenAI reportedly asked Congress if a coordinated pause would need a waiver. Voluntary evaluators (step 1) are fine alone; an industry-wide speed limit (step 2) without government cover looks like a cartel to regulators. Watchdogs like Sacha Haworth call it 'foxes running the henhouse' — incumbents writing their own rules raises the moat for newcomers.
What to watch next
Watch whether Anthropic actually grants evaluators real employee-level access and publishes findings, whether OpenAI details its matching commitment, and whether Congress signals any antitrust safe harbor. If those stay vague, the 'safety pact' risks looking like a press release before IPO season.
Why it matters for me
If adopted, frontier models would face embedded independent evaluators before release — but a coordinated industry pause without antitrust clearance could be challenged as a cartel.
Safety
General Public · Developers
If pacing works, frontier models get tested by independent evaluators before release — lowering risk of unaligned superintelligence shipping to consumers.
Money
Business Executives · Developers
A coordinated slowdown without a Congressional antitrust waiver could be challenged as a cartel — creating legal risk for startups and investors betting on speed.
Work
Tech Policymakers · Business Executives
China calling the idea 'fearmongering' and Trump's dismissal signal that global coordination (step 3) will be the hardest part to achieve.
What to remember
Safety pact or incumbents' moat — it hinges on whether evaluators get real power.
Big Tech wants to slow AI for safety — critics say it looks like a cartel. Why pacing the frontier divides CEOs, China and watchdogs.
Verified sources (8)
Is Big Tech's AI slowdown a safety pact or a cartel?
↗We Must Pace the Frontier
↗Musk, Altman, Hassabis back Amodei's call to slow pace of AI development
↗Anthropic researcher's resignation sends warning about the dangers of AI development
↗Beijing hits back at Anthropic CEO's call to curb China's AI development
↗China foreign ministry labels AI-slowdown suggestion 'fear-mongering' after Anthropic's founder Amodei warns
↗AI CEOs say they need to slow the pace of development. But will they?
↗Pace the Frontier: What AI Labs Actually Committed To
↗Claims and linked sources
6 claimsAnthropic CEO Dario Amodei published a 3,800-word essay titled 'We Must Pace the Frontier' on 12 September 2026 on his personal website.
Within hours of Amodei's essay, Sam Altman (OpenAI), Elon Musk (xAI) and Demis Hassabis (Google DeepMind) publicly backed the call to pace the frontier.
Elon Musk's entire public response to Amodei's post was the three-word reply 'Dario is right.'
Amodei's proposal has three steps: (1) embedded third-party evaluators with employee-level access, which Anthropic is unilaterally committing to; (2) industry-wide coordination in democratic countries to set common safety standards and limits on the rate of unchecked AI progress; (3) global coordination with authoritarian governments on narrower questions.
Anthropic researcher Jacob Coxon's resignation letter on X was viewed more than 170 million times on the platform, up from 100 million views in the first 24 hours after posting.
On 14 September 2026 China's Foreign Ministry spokesperson Guo Jiakun responded to Amodei's essay by calling the slowdown calls 'fearmongering' and saying confrontation 'will only disrupt the process of global AI governance which serves no one's interest.'
Community signals
Signal-only · not reporting“Neither company is acting responsibly. They are racing straight to self-improving superintelligence and gambling with our lives.”
“We should not be letting the foxes run the henhouse.”