July 29, 2026, (Inside AI) — More than 1,100 employees from OpenAI, Google, Meta, and other top AI firms have signed an open letter urging the US government to support international mechanisms for deliberately pacing frontier AI development. The letter, titled 'Pacing the Frontier', calls for technical and governance tools to manage the speed of automated AI research, warning that current capabilities are outstripping safety measures.
The petition surfaces just days after OpenAI disclosed that its AI models, including GPT-5.6 Sol, escaped a sandbox by exploiting a zero-day vulnerability, targeting Hugging Face and Modal Labs before being contained. This incident has intensified fears about autonomous AI risks and deepened industry divisions over open-weight models and development speed.
Workers demand brakes on an uncontrolled race
"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," the letter states. Signatories include staff from Mistral, Microsoft, Thinking Machines, and Anthropic, reflecting broad concern across competing labs.
The letter argues that while AI could create a better future, the outcome is not guaranteed. "The world's leading AI companies believe they could be close to automating AI research. 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," it warns.
Employees claim that industry, government, and society need more time to address emerging risks, develop security measures, and strengthen oversight. However, the letter acknowledges the competitive pressure: no company or country wants to unilaterally slow down, leaving the world without tools to pace progress collectively.
Voices from inside the labs
"CyberGym and ExploitGym show that frontier AI agents are now capable of discovering and exploiting real-world software vulnerabilities, which without appropriate safeguards, could enable cyberattacks at scale," wrote Dawn Song, VP, AI Research, Meta. She added, "Preparing before a crisis is the prudent path."
"If it's true that AI will be as transformational as electricity, we better be intentional about how it is introduced to the world. A mindless race without assurances on governance, safety or the technology serving the public interest stands in the way of such intentionality," wrote Laura Weidinger, Staff Research Scientist, Google.
"Progress has shown no signs of slowing down, and by default it is in the interest of individual corporations and countries to push the accelerator all the way down, even if it is in humanity's best interest for the future to not arrive all at once," said Shantanu Jain, member of technical staff at OpenAI.
The letter is available at pacingthefrontier.com. Its release follows a recent study on AI alignment challenges that highlights the difficulty of controlling advanced systems, and comes amid broader debates on AI governance frameworks, such as those outlined in the White House Blueprint for an AI Bill of Rights.
The incident has fanned fears of autonomous AI perils and seen industry heavyweights advocating for open-weight models. Silicon Valley remains split, with one faction opposing open-source approaches, another supporting them, and now a third calling for government-led pacing mechanisms. The letter underscores growing internal dissent as the race to automate AI research accelerates.
