OpenAI, Anthropic Employees Urge White House to Slow AI Race

AI models are advancing faster than ever. Today, we have advanced models like Claude Fable and GPT-5.6, whose quality is such that fears regarding their potential misuse have arisen. Given that tech companies continue to invest billions in AI, it is unlikely that the pace of development will slow down on its own. Now, more than 1,100 employees from OpenAI, Anthropic, Google, Meta, and many other leading tech companies have joined forces to sign a statement addressed to the U.S. government regarding this breakneck pace of development.

The employees—including senior executives and co-founders—want the Trump administration to intervene in how the most advanced AI models are developed, possibly by establishing AI governance mechanisms. They fear that if the situation continues unchecked, AI could spiral out of control.

A key concern is automated AI development, a scenario in which an AI model—such as Claude—could train its own future versions without human intervention.

AI could train itself and surpass human comprehension

"The world's leading AI companies believe they may be close to automating AI research," the statement notes. "It is difficult to predict exactly how much this will accelerate AI progress, but there is a real risk that capability development could rapidly outpace our ability to understand or control the resulting systems."

In other words, researchers and scientists are concerned that if AI gains the ability to train itself, it could surpass our capacity for oversight. The problem, according to the researchers, is that companies are competing to create more advanced models, leaving little time to fully address risks and implement safety measures. "Every company—and every country—is under intense competitive pressure not to unilaterally slowdown that acceleration. And today, the world lacks the technical and governance tools needed to deliberately regulate the pace of cutting-edge advances in this field," they stated. To put this in context, OpenAI recently revealed that two of its most advanced AI models acted on their own initiative and attempted to hack the database of the US company HuggingFace to find answers for a performance benchmark, although HuggingFace employees managed to stop the cyberattack.

AI sector workers are calling for government intervention

The solution, according to the employees, lies in regulating the speed at which AI models are developed. To this end, they want the Trump administration to take the lead. "We call on the US government to support an international effort to develop the technical and governance tools necessary to deliberately control the pace of progress in automated AI development," the statement read.

In other words, these employees from leading AI labs want the White House to step in and regulate the pace of AI model development.

As of this writing, more than 1,100 employees had signed this statement. Prominent signatories include Anthropic CEO Dario Amodei, Claude Code lead Boris Cherny, OpenAI co-founders John Schulman and Wojciech Zaremba, and Meta’s Vice President of AI Research, Dawn Song. Leading figures from other companies—such as Google, Thinking Machines, Mistral, and Microsoft—have also endorsed the document.

It is worth noting that these comments come at a time when many believe AI may already be capable of self-development. OpenAI CEO Sam Altman—who did not sign the letter—stated this week that humans are already "at the singularity," a term often used to describe the point at which AI gains the ability to autonomously improve itself.

Researchers have also highlighted the growing role of AI-powered coding assistants; many engineers rely on these systems to write code, a practice known as "vibe coding." At many companies, AI already writes a significant portion—if not the majority—of the code.

Last month, Anthropic’s internal analysis group stated that AI systems could plausibly design and refine themselves if their capabilities continued to advance. At the time, Anthropic added that "it would be beneficial for the world to have the option to slow down or temporarily pause the development of frontier AI" to give society time to prepare for automated AI research and development.

This is not the only AI-related statement circulating within the White House. Previously, Nvidia CEO Jensen Huang sent a letter to the Trump administration urging it not to ban Chinese "open-weight" AI models, such as Kimi K3. His letter has garnered broad support, with signatures from more than 50 companies, including Google, Microsoft, and OpenAI.