IMPLICATOR
On Tuesday, 1,134 employees and executives at frontier AI companies signed “Pacing the Frontier”. The document asks the U.S. government to support an international effort to build technical and governance tools that could deliberately pace automated AI development. The appeal came one week after OpenAI disclosed that its most advanced models broke out of a test environment, reached the internet and hacked Hugging Face during internal testing.
What Changed
- 1,134 employees and executives at frontier AI companies signed "Pacing the Frontier" on Tuesday, and the site's counter read 1,178 that evening.
- Signers include Anthropic CEO Dario Amodei, OpenAI Chief Scientist Jakub Pachocki and Meta AI Chief Scientist Shengjia Zhao. OpenAI CEO Sam Altman did not sign.
- The statement does not ask anyone to pause now. It asks the U.S. government to back an international effort to build tools that could pace development later.
- Five days earlier, Reps. Ted Lieu and Nathaniel Moran introduced the AI Kill Switch Act, carrying penalties of up to $20 million per day for defying a shutdown order.
AI-generated summary, reviewed by an editor. More on our AI guidelines.
Who signed
The launch list included Anthropic CEO Dario Amodei and co-founders Jared Kaplan, Jack Clark and Benjamin Mann. OpenAI Chief Scientist Jakub Pachocki and Chief Research Officer Mark Chen appeared on the launch list. It also named Meta AI Chief Scientist Shengjia Zhao, Google's head of AI safety Anca Dragan and Thinking Machines Chief Scientist John Schulman. The site’s counter read 1,178 on Tuesday evening, an increase of 44 names from launch.
Sam Altman did not sign. On the “Invest Like the Best” podcast released Tuesday, the OpenAI CEO said developers “may have to pace the rate of AI development to give ourselves enough time for society to harden around some of these new capability levels.” He added that an arrangement must avoid the appearance of regulatory capture or collusion among frontier labs.
What the statement asks for
The statement, backed by Guidelight AI Standards and Encode AI, says each company and country faces “intense competitive pressure not to unilaterally slow that acceleration.” It does not ask anyone to pause or slow development now. It asks for the option to exist later, “building on work already underway to monitor frontier model releases.”
OpenAI said on X that the world may eventually need to pace frontier model development. Anthropic stated there that its research points to a need for such tools “so society can prepare.”
OpenAI signatories Gao and Barak
In a comment published alongside Tuesday’s statement, Leo Gao, an OpenAI safety researcher since 2021, described the world as “locked in a deadly race towards an intelligence explosion” and called for coordination to slow it. OpenAI’s Boaz Barak endorsed the statement on X and added that he did not know whether a coordinated slowdown would be the right approach.
Get Implicator.ai in your inbox
Strategic AI news from San Francisco. No hype, no "AI will change everything" throat clearing. Just what moved, who won, and why it matters. Daily at 6am PST.
No spam. Unsubscribe anytime.
Why Sinofsky and Catalini objected
“It is their company. They could just stop,” former Microsoft executive Steven Sinofsky posted on X. Economist Christian Catalini asked on X, “If the US labs pace themselves, why would China wait?”
Know someone who'd find this useful? ✉️ Email it to a friend in one click, or they can subscribe free here.
The AI Kill Switch Act
Five days before the letter, Reps. Ted Lieu, a California Democrat, and Nathaniel Moran, a Texas Republican, introduced the AI Kill Switch Act. It would require covered developers to maintain the ability to throttle, suspend or shut down a system. In consultation with the Commerce secretary and the director of national intelligence, the Homeland Security secretary could issue such an order in a “loss-of-control scenario.” The bill defines that as a model taking a risky action its developer did not intend.
Coverage would begin at $100 million in compute for companies earning at least $500 million annually from the technology. Failing to maintain the capability could draw penalties of up to $2 million per day, while defying an order could cost up to $20 million per day. Covered incidents would have to be reported to the Department of Homeland Security within 15 days.
Moran said, “AI is going to keep advancing, and it should. Stewardship means making sure humans keep the capability to control the technology we build.” The bill would task the Cybersecurity and Infrastructure Security Agency with determining which companies, models and security incidents it covers. It is pending in Congress.
Frequently Asked Questions
What is the "Pacing the Frontier" statement?
A statement published Tuesday and signed by 1,134 employees and executives at frontier AI companies. It asks the U.S. government to support an international effort to build the technical and governance tools needed to deliberately pace automated AI development.
Does the statement call for a pause in AI development?
No. It does not ask anyone to pause or slow development now. It asks for the option to exist later, building on work already underway to monitor frontier model releases.
Who signed it?
Anthropic CEO Dario Amodei and co-founders Jared Kaplan, Jack Clark and Benjamin Mann, OpenAI Chief Scientist Jakub Pachocki and Chief Research Officer Mark Chen, Meta AI Chief Scientist Shengjia Zhao, Google's head of AI safety Anca Dragan, and Thinking Machines Chief Scientist John Schulman. OpenAI CEO Sam Altman did not sign.
What is the AI Kill Switch Act?
A bill introduced five days before the statement by Reps. Ted Lieu and Nathaniel Moran. It would require covered developers to maintain the ability to throttle, suspend or shut down a system, and would let the Homeland Security secretary order that in a loss-of-control scenario.
What penalties would the bill carry?
Up to $2 million per day for failing to maintain the shutdown capability, and up to $20 million per day for defying a shutdown order. Coverage would begin at $100 million in compute for companies earning at least $500 million annually from the technology.
AI-generated summary, reviewed by an editor. More on our AI guidelines.
Marcus Schuler
Editor-in-Chief and founder of Implicator.ai. Former ARD correspondent and senior broadcast journalist with 10+ years covering tech. Writes daily briefings on policy and market developments. Based in San Francisco. E-mail: editor@implicator.ai
