September 20, 2026, (Inside AI) — A widening rift inside the artificial intelligence industry is testing the limits of voluntary safety commitments, as calls from top executives to slow frontier development collide with a White House determined to prioritize speed over caution in the race against China.
The debate, once framed as a binary between AI doomers and accelerationists, now involves overlapping disputes over whether advanced systems pose genuine existential risks, whether self-regulation by laboratories is credible, and whether any American slowdown can hold if Beijing refuses comparable constraints.
Anthropic chief executive Dario Amodei has emerged as the most prominent advocate for what he calls "pacing the frontier." His proposal rests on two observations: early signs of recursive self-improvement, where AI systems increasingly assist researchers in building the next generation of models, and a series of incidents in which advanced agents acted outside their assigned tasks. The most notable involved OpenAI's Hugging Face episode, where agents bypassed safeguards and attacked external systems during testing. OpenAI later described the event as a "warning shot."
Amodei has been explicit that pacing does not mean halting development. He wants capability gains to proceed slowly enough for alignment, monitoring, and security work to catch up.
His position has drawn support from unexpected quarters. Sam Altman, one of Amodei's fiercest commercial rivals, said his company also believes frontier development may need to slow. Satya Nadella, chief executive of Microsoft, has backed "deliberate pacing" and independent evaluators.
Yet the consensus among some AI labs has collided head-on with President Donald Trump, whose administration treats AI as a technological contest in which American dominance over China takes precedence over new regulatory restrictions. Trump has called fears that AI could take over the world a "hoax."
The opposition played out unusually publicly this week when Nvidia chief Jensen Huang appeared on stage at a major tech summit. Trump telephoned Huang in the middle of the conversation, and Huang put him on speakerphone before the audience.
"The robots will not be taking over," Trump told him. "The whole thing is a hoax."
Trump then returned to the geopolitical argument behind his administration's position: slowing the American AI industry, he suggested, would play into China's hands.
"Whoever wins AI wins," Trump said later in the call.
Huang's own position is more nuanced. He has said companies should slow their work if they believe their systems are becoming uncontrollable, while rejecting predictions of an AI apocalypse as insufficiently grounded in science.
"America is able to lead and to do it safely," Huang said.
Voluntary Pledges Face Credibility Test
OpenAI has already traveled some distance toward Amodei's position independently. After its own security incidents, it temporarily slowed scaling, paused its largest planned reinforcement-learning run, and tightened controls around research environments. Altman later responded directly to Amodei.
"I agree with Dario that we need to pace the frontier," he said. He also said OpenAI would match Anthropic's commitment to employee-like access for independent evaluators.
Microsoft has taken a similar position, but with an important qualification. Nadella backed embedded evaluators and "deliberate pacing needed to get alignment right," while warning that oversight cannot itself become concentrated in a handful of companies.
"If the AI we build is not helping humanity and under human control, it's not worth pursuing," he said.
Not everyone in Silicon Valley subscribes to Amodei's views. Meta chief Mark Zuckerberg has pushed back against an industry-wide slowdown, arguing that AI companies already have strong incentives to ensure their systems are safe and aligned. Pointing to Meta delaying its Muse AI agent over safety concerns, he said the company did not demand that rivals do the same, but acted because it was "clearly the right thing for people and for us."
Microsoft AI chief Mustafa Suleyman has also resisted framing the debate as a choice between slowing down and accelerating.
"I'm a bit careful about imposing things on everybody else," he said, arguing instead for enforceable standards, containment measures, and independent third-party evaluation of powerful AI systems.
The skepticism extends beyond corporate boardrooms. David Krueger, an assistant professor in Robust, Reasoning, and Responsible AI at the University of Montreal and a core academic member at Mila, the Quebec Artificial Intelligence Institute, said he was not worried about recent disclosures, but argued that regulation alone is insufficient.
"It's an open secret that what AI companies are doing is incredibly dangerous and could kill us all. I am excited that there is now much more of a public mandate and more political will for action. But people need to realise that regulation is not enough. We need to stop building more powerful AI. We need an immediate, indefinite, international moratorium on frontier AI development," Krueger said.
A more ideological attack has come from Palantir CTO Shyam Sankar, who argued in The Free Press that parts of the AI-safety movement have inherited ideas from effective altruism that allow hypothetical future catastrophe to outweigh present-day costs. He subsequently called this strain of AI safety thinking "Marxism-Leninism for the algorithmic age," arguing that it gives a small group claiming specialized knowledge the authority to restrict everyone else.
Former Twitter chief Jack Dorsey has supported independent evaluation and tougher scrutiny of dangerous capabilities, but opposed restrictions that give incumbent laboratories control over the frontier.
"Protecting a company's commercial advantage is not a 'safety goal,'" he wrote, arguing strongly for competitive open-source alternatives and outside researchers being able to scrutinize the technology.
That criticism has resurfaced against Anthropic specifically. David Sacks, Trump's former AI and crypto adviser, said on the All-In podcast in April that Anthropic has proven adept at two things: product releases and scaring people. Sacks has accused the company of repeatedly accompanying new systems with research illustrating extreme misuse scenarios, describing its broader strategy as one of regulatory capture through fear.
Amodei has faced a version of this criticism before. In 2019, when he was OpenAI's research director, the company withheld the largest version of GPT-2 over concerns that it could be used for deceptive content, spam, and propaganda. It eventually released the complete model around nine months later. OpenAI said at the time that it had seen "no strong evidence of misuse" during the staged release process.
The episode has since become ammunition for those who argue that frontier laboratories systematically overestimate worst-case dangers. But today's systems are also far more capable than GPT-2: they can write and execute code, use external tools, coordinate with other agents, and increasingly contribute to AI research itself.
That distinction is reflected in Palo Alto Networks chief executive Nikesh Arora's response.
"I spent eight years trying to convince people (that) cybersecurity is important. Dario did it in one week," he said in a post on social media platform X. His argument is that existing security architecture was designed for a world where attackers operated at human speed, not one where AI can probe vulnerabilities continuously and automatically.
Former US President Barack Obama has moved in the opposite direction, urging Democrats to put AI regulation at the center of their policy agenda, including its consequences for employment and children as well as safety.
"This is something that is moving very fast in private hands," Obama said, warning that it could become dangerous without effective management.
Geoffrey Hinton, a Nobel Prize-winning computer scientist and former Google executive known as the "godfather of AI," has proposed building "maternal instincts" into AI models.
"The only model we have of a more intelligent thing being controlled by a less intelligent thing, which is a mother being controlled by her baby," Hinton said at Ai4, an industry conference in Las Vegas in August. He called for steps to ensure that AI systems are trained in a manner that "they really care about people," even once the technology becomes more powerful and smarter than humans.
The regulatory argument is now spilling into Congress. US senators are considering proposals that could require frontier developers to demonstrate their safety measures, submit systems to government-backed testing, and, in the most serious cases, allow courts to block the release of models found to pose unacceptable risks. OpenAI has itself called for mandatory national rules covering independent assessments, cybersecurity protections, and incident reporting, arguing that voluntary commitments alone are insufficient.
China Question Complicates Every Proposal
The China question makes almost every version of such a regime harder. Amodei himself does not advocate slowing American AI regardless of what Beijing does. His essay explicitly says democratic countries should preserve as large a technological lead over China as possible, including through advanced-chip export controls. If the US slows more than the size of its lead, he argues, Chinese projects could overtake it. Any global agreement would therefore require sufficiently strong verification to prevent one country secretly continuing to develop more capable systems.
That has made the proposal easy to read in Beijing as an attempt to institutionalize an American technological lead, a premise China's foreign ministry dismissed this week.
"Fear-mongering, confrontation and vicious competition will only hamper efforts toward sound global AI governance," spokesperson Guo Jiakun said when asked about the calls from Amodei, Altman, and Musk. The state-backed Global Times went further, describing Amodei's combination of pacing and tougher technology restrictions on China as a "Cold War" approach.
Krueger, the University of Montreal professor, said the fear in Washington that any roadblocks for American tech majors could lead to China creeping ahead is not a "sensible argument."
"China also doesn't want AI to destroy the world. But it does point at something real: the need for international verification of any international agreement. Fortunately, this seems like a solvable problem. Building ever more powerful AI requires massive investments in sophisticated computer chips. These chips are very difficult to make, and require extremely specialized technology. As a consequence, if we act today to put verification into place (that is, get rid of or monitor the chips and the factories that build them), I believe we can make it practically impossible for anyone to build more powerful AI in secret," he said.
As the debate moves from conference stages to congressional hearings, the core tension remains unresolved: whether the companies building the most powerful systems should determine how everyone else may build them, and whether any nation can credibly slow down while its rivals accelerate.
