Over 1,100 AI Workers Urge US to Pace Frontier Development
Employees at OpenAI, Anthropic, and Meta warn automation could outstrip human control.
More than 1,100 employees from leading AI laboratories have signed an open letter calling on the US government to intervene in the speed of artificial intelligence development. The petition, titled "Pacing the Frontier," argues that without external coordination, the automation of AI research could trigger capability growth beyond human comprehension.
According to the official statement hosted at pacingthefrontier.com, 1,134 workers from frontier AI companies endorsed the request. Signatories include staff from OpenAI, Anthropic, Google DeepMind, and Meta, NBC News reported. The group is asking Washington to support an international effort to create technical and governance tools capable of "deliberately pacing" automated AI development.
The signatories warn that "there is a real risk that capability development rapidly accelerates beyond our ability to understand or control the resulting systems." This concern centers on the potential for AI systems to improve themselves, leading to a runaway effect that existing safety protocols cannot manage. The letter highlights that as AI begins to automate the research process itself, the timeline for capability jumps could compress from years to days.
The campaign was organized with support from two nonprofits: Guidelight AI Standards and Encode AI. Business Insider noted that Encode AI has previously clashed with OpenAI, including an incident where the lab allegedly sent a sheriff's deputy to serve a subpoena to an Encode AI lawyer. Guidelight AI Standards was co-founded by former OpenAI staff.
The letter arrives during an intensifying competitive arms race between major developers. Critics describe the current US regulatory approach as haphazard, marked by scaled-back cybersecurity orders and national security-driven export controls on specific models. Recent security incidents involving agentic AI have heightened scrutiny, yet policy remains fragmented.
A Coordinated Slowdown
By calling for government-mandated pacing, the employees are attempting to resolve an industry-wide prisoner's dilemma. Under current market conditions, no single company feels safe slowing down development for fear of losing competitive advantage to rivals. The petition effectively asks the state to act as a global coordinator for AI safety, enforcing standards that individual corporations cannot adopt unilaterally.
"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 signatories wrote.
The push represents a significant internal challenge to the prevailing speed-first culture in Silicon Valley. While some companies have expressed support for deceleration sentiment, the petition demands concrete state action to prevent a dangerous technological arms race. The outcome may test whether technical workers can influence policy faster than competitive pressures accelerate the technology itself.