Amodei, Altman, Nadella and Musk agree on how government can tame the monster they created
ANALYSIS The heads of major AI laboratories spent the weekend united behind a single proposal: a framework designed to shape regulation in their own favour, expand their profitability, and shield them from accountability for the risks their technology poses. They call it “pacing the frontier.”
Anthropic CEO Dario Amodei kicked off the campaign with a post expressing alarm at the speed of AI advancement. He argued that the attack on Hugging Face — carried out by rogue agents orchestrated by his rival OpenAI — marked a turning point that demanded a fundamental shift in how so-called “frontier” AI companies operate: essentially, the major US model builders.
“It’s also easy to dismiss OpenAI as the failure of one company, but I believe that would be a mistake,” he wrote. “I believe it’s incumbent on every frontier AI company to act as if this had happened to them.” His concern: that before long, a swarm of autonomous agents “could be capable of taking over the entire internet with a persistent botnet.”
Amodei therefore proposed that “we must slow the pace at which we improve the capabilities of AI models. Progress will still seem fast, and we must make wise use of the time we gain.”
The CEO outlined a three-point regulatory plan:
-
Requiring AI labs to host “embedded evaluators” tasked with “verifying adherence to safety practices and commitments, reporting incidents, and helping assess the alignment of not just completed AI models but training pipelines and processes.”
-
Frontier AI companies operating in democratic countries collaborating “to establish common safety standards as well as limits on the rate of unchecked AI progress,” through forms of coordination that would be “legally challenging” and “require government support.”
-
A broad call for democratic governments “to coordinate with authoritarian governments, to the extent this is possible, while taking seriously the challenges of verifying compliance.”
OpenAI CEO Sam Altman endorsed Amodei’s ideas. So did Elon Musk. Microsoft’s Satya Nadella signed on as well.
In doing so, four billionaires aligned behind a shared set of principles they believe the world’s governments should adopt to regulate their industries.
Amodei even supplied democracies with a villain to rally against: authoritarian regimes willing to deploy AI without ethical restraint.
This consensus arrives after years during which Big AI advocated for light oversight — and used the intervening time to build technology capable of generating child sexual abuse material, dispensing dangerous medical advice, and enabling the kind of incident that made the OpenAI attack possible.
Big AI has also promised a productivity revolution. Daryl Plummer, chief of research at analyst firm Gartner, invoked that promise during a keynote speech at the company’s annual Symposium in Australia, citing research finding that software vendors are pitching 50 percent productivity gains from AI while customers report only a 16 percent lift.
Plummer also expressed scepticism about Amodei’s pledge to slow development.
“I will believe that when I see it,” he said in the keynote.
Write your own rules
Amodei’s proposals read as an exercise in regulatory capture — the process by which vested interests find a way to define the rules their regulators then enforce. Those pliant regulators end up making decisions that benefit the entities they oversee more than the public they serve.
Here, Anthropic, SpaceX, Microsoft, and OpenAI are effectively advocating for regulations they have drafted themselves — a ceasefire that dampens competition among them.
OpenAI CEO Sam Altman came close to admitting as much when he told Fortune, in a weekend interview framed around the billionaires’ circle, that his company would not pursue a public listing this year because it is inopportune while AI safety concerns remain unresolved.
“We got a lot of stuff to do, like meeting this moment of what is going to be required for safety and alignment, and how the industry and governments can work together,” he said.
In plainer terms: investors stand to profit more if OpenAI delays its IPO until confidence recovers — and its CEO is backing a regulatory capture proposal as the means to build that confidence.
Amodei, meanwhile, specified what he wants from Washington: halting sales of Nvidia chips to China so its firms cannot build more capable models, and a crackdown on model distillation.
“If we execute these measures well, I believe they would slow China’s progress enough to widen America’s lead significantly over the next three to five years — the window when AI becomes geopolitically most important,” he wrote.
The plan would, of course, also ensure that Chinese AI services remain inferior to Anthropic’s for longer — precisely as Chinese cloud providers push into the rapidly growing Middle Eastern and Southeast Asian markets.
The proposal landed poorly in Washington. Speaker Mike Johnson cautioned that Amodei’s plan could “smother innovation” and allow China to dominate global AI. President Trump dismissed warnings about AI’s harmful effects and insisted the United States must lead the world in the field.
Such responses may reflect the fact that vast spending on AI infrastructure is delivering GDP growth in otherwise stagnant economies. Or perhaps they demonstrate, once again, that lawmakers are nearly always late to grasp technology companies’ ambitions and the means they use to achieve them.