.png)
In Brief
More than 1,300 employees at top AI companies such as Meta, Google, OpenAI and Anthropic asked the US government to help pace AI that automates AI research.
Signers include some of the field's top scientists.
What Happened
More than 1,300 employees across nearly a dozen top AI companies, including OpenAI, Anthropic, Google, and Meta, signed a statement called Pacing the Frontier. The core request is one sentence: "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."
The signers are not fringe voices: Anthropic CEO Dario Amodei, OpenAI Chief Scientist Jakub Pachocki, Safe Superintelligence CEO Ilya Sutskever, Google DeepMind co-founder Shane Legg, and Meta Superintelligence Labs Chief Scientist Shengjia Zhao. Both OpenAI and Anthropic backed it with official statements.
SmarterX founder and CEO Paul Roetzer unpacked it on Episode 228 of The Artificial Intelligence Show.
The Key Numbers
1,300+ - Frontier AI company employees who signed the Pacing the Frontier petition
2028 - OpenAI's internal target for a significant fraction of its research done by AI systems
Factor of five - Cost threshold in Anthropic's policy for models fully substituting for its research staff
2x - Rate-of-progress jump that triggers Anthropic's automated R&D capability threshold
Why No Lab Can Slow Down on Its Own
The pace message is everywhere. The petition follows Google DeepMind CEO Demis Hassabis' essay calling AGI (Artificial General Intelligence) "much more akin to the discovery of electricity or fire" and proposing a standards body. AGI is AI that can do most economically valuable work as well as a human. Bloomberg reported that OpenAI CEO Sam Altman told reporters in Washington, "We've talked about the need to pace it as the models get more capable, which I think is in everyone's interest." Of the petition, Altman said, "We helped participate in the language on that."
OpenAI is asking for a brake on its own goal. The frontier lab's plan, "built to benefit everyone," lists this goal: "Build an automated AI researcher, an AI system that can accelerate and increasingly automate the research process itself while remaining steerable, accountable, and connected to people." OpenAI believes that by March 2028, a significant fraction of its research may be done by AI systems. "You're asking the government to save you from yourself," Roetzer says. "We're going to achieve this, but you might want to slow us down. We can't slow ourselves down."
Anthropic has drawn the red line. Its Responsible Scaling Policy sets the trigger: its models fully substituting for its entire research staff at competitive costs, within a factor of five, a dramatic acceleration in the pace of AI progress, which Anthropic would measure as a doubling of the rate of progress likely attributable to automated AI research. "All of these thirteen hundred plus researchers are seeing a trend line that tells them they are moving faster in the advancements of automated AI research than they are comfortable with," Roetzer says.
"We may be one turn away on frontier models to where these [researchers] no longer feel comfortable that they fully understand what the models are capable of and how to put guardrails in place to safely release them into the world."
— Paul Roetzer, founder and CEO of SmarterX, Episode 228 of The Artificial Intelligence Show
Stopping alone looks like losing everything. If Anthropic stopped at its threshold while a Chinese lab or rival kept going, it believes its models would no longer be sufficient to protect it. Once a rival's model begins recursive self-improvement or AI improving itself and making itself better with each research cycle, the lead becomes insurmountable.
SmarterX Take
Critics call what the labs are doing regulatory capture, when big labs seek rules that lock out open models and rivals. "The people who just keep screaming regulatory capture as though that one argument is the simple reason why everybody feels this way and they're calling for this, I think it's just doing a disservice to the industry to think that that's all that's happening here," Roetzer says.
"It's just too much evidence in the other direction that we are truly entering a dangerous realm here that they do not feel comfortable with where these models are going and their ability to control them."
What to Watch
Whether Washington builds anything. The tools the petition asks for do not exist today.
Meta's moves. Meta CEO Mark Zuckerberg's same-day Wall Street Journal op-ed argues the defining question is not whether superintelligence arrives but who gets to use it. Roetzer noted Meta has not voluntarily submitted its models for evaluation.
What 2,100+ Professionals Expect AI to Do to Jobs
Seventy-one percent of professionals believe AI will eliminate more jobs than it creates over the next three years, according to the 2026 State of AI for Business Report. Only 13% expect net job creation. The people building frontier AI are asking for time to keep it under control, and the workforce is already bracing for its impact.
The report is built on more than 2,100 responses across roles, functions, and industries, and it maps how professionals and organizations are actually adapting to AI right now. If the insiders are worried about pace, the smartest response for business leaders is to understand exactly where their own people stand. Read the full report →
Mike Kaput
Mike Kaput is the Chief Content Officer at SmarterX and a leading voice on the application of AI in business. He is the co-author of Marketing Artificial Intelligence and co-host of The Artificial Intelligence Show podcast.