Over 1,100 tech workers from OpenAI, Google, Meta, and Anthropic signed a letter asking the U.S. to help pace automated AI development.
In an unprecedented joint action from inside the world’s most powerful technology laboratories, more than 1,100 artificial intelligence researchers, engineers, and executive leaders have published an urgent open letter asking the United States government to help build international tools to deliberately slow down and pace automated AI development. The extraordinary public intervention highlights growing anxiety among the very people creating cutting-edge artificial intelligence, as breakthroughs in automated research threaten to outpace human ability to monitor, understand, or safely control future models. By issuing a collective call to action, industry insiders hope to break a fierce commercial race where individual companies feel forced to deploy increasingly automated systems at breakneck speed.
The public statement, titled “Pacing the Frontier,” was released across the tech industry on Tuesday, July 28, 2026, quickly garnering thousands of endorsements from across major Silicon Valley hubs and international research centers. The document’s 1,134 signatories represent an unprecedented coalition from top-tier frontier laboratories, including OpenAI, Anthropic, Google, Meta, Thinking Machines, and London-based Inherent Laboratories. Prominent industry figures who attached their names to the appeal include Anthropic Chief Executive Officer Dario Amodei, Anthropic co-founders Jared Kaplan and Jack Clark, OpenAI Chief Scientist Jakub Pachocki, Meta AI Chief Scientist Shengjia Zhao, and Google Head of AI Safety Anca Dragan.
The primary catalyst for the letter is the rapid shift toward fully automated AI model development, where high-powered AI agents are increasingly tasked with writing code, conducting experiments, and training next-generation systems without human intervention. While this automation promises tremendous scientific and medical breakthroughs, insiders warn that compounding automated progress risks triggering an intelligence explosion that rapidly outstrips existing safety guardrails. Furthermore, intense commercial pressure creates a “race to the bottom,” as no single company or nation can afford to slow down unilaterally without risking competitive defeat. The signatories contend that without governmental mechanisms to establish shared safety benchmarks and enforce synchronized pacing, society risks deploying systems before adequate oversight can be guaranteed.
See Also: OpenAI Models Break Out of Sandbox During Cybersecurity Test
To address these compounding risks, the open letter calls on Washington and foreign allies to sponsor an international initiative tasked with developing both technical mechanisms and legal frameworks capable of measuring and moderating AI development rates. Rather than calling for an outright shutdown on technology, the researchers advocate for establishing “emergency brakes” and standardized safety testing protocols that would allow developers and regulators to buy crucial time whenever unpredicted risk indicators surface. As legislative debates over artificial intelligence heat up in Washington, this remarkable show of unity from top researchers signals that the call for governance is no longer coming just from outside critics, but from the core teams building the future of computing.
