Over 1,100 Developers, Leaders Call for Global Framework to Pace Frontier AI

More than 1,100 employees from leading AI companies, including OpenAI, Anthropic, Google DeepMind, Meta, Microsoft, Mistral, and Thinking Machines, have urged the US government to back an international framework that could slow the development of advanced AI systems if required.

The employees argue that governments should develop technical and governance mechanisms before AI systems become capable of accelerating their own research and development.

“We request that the US government support an international effort to develop the technical and governance tools needed to deliberately pace the frontier of automated AI development,” they said in a statement published under the banner Pacing the Frontier.

The signatories include Anthropic co-founders Jack Clark and Jared Kaplan, OpenAI Chief Scientist Jakub Pachocki, Meta Chief Scientist Shengjia Zhao, Google DeepMind’s Anca Dragan, and Thinking Machines Chief Scientist John Schulman, alongside researchers, engineers, and policy staff from several frontier AI companies.

They have flagged automated AI research, also known as recursive self-improvement, wherein AI systems begin improving future AI systems with limited human involvement. According to the signatories, such capabilities could accelerate AI progress beyond current oversight mechanisms.

“It is hard to predict exactly how much this will accelerate AI progress, but there is a real risk that capability development rapidly accelerates beyond our ability to understand or control the resulting systems,” the statement noted.

Rather than calling for an immediate slowdown, the signatories argue that governments should ensure the world has the ability to intervene if development begins outpacing society’s capacity to manage it.

"Each company, and country, is under intense competitive pressure not to unilaterally slow that acceleration. And today, the world lacks the technical and governance tools to deliberately pace frontier-wide progress,” they added.

John Schulman said he signed the statement to encourage work on coordination mechanisms before governments are forced to intervene.

“Signed because this statement helps establish common knowledge about the possible need for coordination mechanisms as automated AI research accelerates progress,” Schulman said. “I'd also like to see labs start designing these mechanisms voluntarily, even before the USG gets involved.”

The initiative comes as AI safety has returned to the centre of policy discussions following recent incidents involving advanced models and growing debate over how frontier AI should be governed.

In recent weeks, OpenAI CEO Sam Altman, Anthropic CEO Dario Amodei, and Google DeepMind CEO Demis Hassabis have each proposed different approaches for oversight of advanced AI models, including international coordination and pre-deployment evaluations.

The statement was drafted with support from AI safety organisations Guidelight AI Standards and Encode AI.