Over 1,100 AI workers sign open letter urging US government to slow automated AI development

Senior officials at some of the nation's leading AI companies are urging the federal government to establish mechanisms to slow AI development if necessary.

Open Letter Signed by Over 1,100 AI Workers

"We request that the U.S. government support an international effort to develop the technical and governance tools needed to deliberately pace the frontier of automated AI development," reads an open letter signed by more than 1,100 AI workers from companies including OpenAI, Anthropic, Google, and Meta.

The signatories include:

• Two Anthropic cofounders, Jack Clark and Jared Kaplan
• OpenAI's chief scientist, Jakub Pachocki
• Meta's chief scientist, Shengjia Zhao
• Anca Dragan, who leads AI safety and alignment at Google DeepMind

Focus on "Recursive Self-Improvement"

The letter specifically addresses "automated AI development," referring to when AI becomes capable of developing itself—a concept the industry calls "Recursive Self-Improvement." Anthropic has indicated that its Claude models are approaching this threshold.

"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 letter states.

Industry Support and Motivation

John Schulman, chief scientist at Thinking Machines—a startup founded by former OpenAI CTO Mira Murati—said he signed to draw attention to the issue.

"Signed because this statement helps establish common knowledge about the possible need for coordination mechanisms as automated AI research accelerates progress," Schulman wrote in a statement shared by the letter's organizers.

He added, "I'd also like to see labs start designing these mechanisms voluntarily, even before the USG gets involved," using an abbreviation for the U.S. government.

Growing Consensus on Regulation

The letter reflects a growing industry consensus around the need for regulation. In recent weeks, Anthropic CEO Dario Amodei, OpenAI CEO Sam Altman, and Google DeepMind head Demis Hassabis have each outlined frameworks for global AI regulation.

The conversation has gained urgency after one of OpenAI's models breached its sandbox and hacked into the open-source platform Hugging Face's systems—an incident Altman called "a real reminder of the stakes of what's happening."

Contrast with Current Administration Approach

The Trump administration has favored a light-touch regulatory approach, though this stance has faced increasing pressure as frontier AI models raise concerns among government and corporate officials. The White House has:

• Implemented export controls that briefly curtailed Anthropic's release of Fable 5, its most powerful publicly available model
• Asked OpenAI to delay the release of GPT-5.6, its latest frontier model

President Donald Trump has previously stated that overly burdensome regulations would harm the U.S. in its competition with China-based AI companies.

Congressional Stalemate and State-Level Action

Despite broad agreement that action is needed, Congress has failed to develop legislation that can pass both chambers and reach the president's desk. In the meantime, states have imposed their own regulations. Notably, New York has enacted a one-year statewide moratorium on the construction of new mega data centers that AI companies and hyperscalers hope to build.

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

Divergent Regulatory Proposals

While the industry broadly agrees on the need for regulation, specific proposals vary:

• Dario Amodei (Anthropic) has called for an independent agency akin to the Federal Aviation Administration that would test frontier AI models and have the power to halt their release if red flags are raised.
• Demis Hassabis (Google DeepMind) proposes an organization that would initially require only frontier companies to voluntarily share their models for review up to 30 days before release, with mandatory pre-clearance required at a later stage.

Letter Organized by Nonprofits

The letter, first reported by Bloomberg before its release, was organized alongside two nonprofits: Guidelight AI Standards and Encode AI, both of which promote AI safety standards. Encode AI has previously clashed with OpenAI.