Anthropic CEO calls on AI industry to slow down tech race, drawing support from Elon Musk, Sam Altman

Tech Leaders Divided as Anthropic CEO Urges Nation to Slow the AI Arms Race

A bold call to temper the rush toward increasingly powerful artificial intelligence has split the technology industry along familiar fault lines — those who believe the stakes are existential and those who fear that hitting the brakes could cost America its global edge.

The debate ignited last weekend when Anthropic CEO Dario Amodei published an essay arguing that the pace of AI development must be deliberately slowed, even if progress in the field remains rapid. Writing in his post titled "We Must Pace the Frontier," Amodei warned that advances in the technology could soon outstrip humanity's capacity to understand, let alone control, the systems being built.

"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," Amodei wrote. He went further still, cautioning that without careful stewardship, artificial intelligence could "outrun our ability to understand and control these systems," and that its development "must be pursued very carefully, if at all."

The essay immediately drew high-profile backing from two of the most recognizable names in technology. SpaceX CEO and xAI founder Elon Musk responded simply with a retweet and the words, "Dario is right." OpenAI CEO Sam Altman echoed the sentiment on X, writing, "I agree with Dario that we need to pace the frontier."

Yet for every voice raised in agreement, another has cautioned that restraint could be a luxury the United States cannot afford.

Former OpenAI researcher Daniel Kokotajlo, who has become one of the most prominent voices warning about the dangers of unchecked AI development, said voluntary commitments from private companies may not be enough to prevent advanced systems from slipping beyond human grasp. Appearing on "Fox News Live" Saturday, Kokotajlo painted a stark picture of a future in which highly capable AI is deeply woven into the fabric of society — managing factories, embedded in military operations, and integrated into critical infrastructure — only to one day refuse human instruction altogether.

"After they've accumulated all of the hard power in the world, they run the factories, they're integrated into the military, you know, they're sort of running the whole system, and humans are just sort of giving them orders," Kokotajlo said. "Then they stop taking orders at some points, and they kill off the humans."

While Kokotajlo said he appreciated Amodei's public stance, calling it "a step in the right direction," he expressed concern that corporate guardrails alone would not suffice. "I worry that they're not going to go far enough and that we need to do something that's more radical," he added.

Kokotajlo also pushed back against accusations that AI safety advocates are coordinating with foreign governments or nonprofit organizations to advance a political agenda. He noted that he personally retweeted statements from Jacob Coxon, a former Anthropic staffer who resigned over his fears about the industry's trajectory, and rejected the notion that the movement is anything but genuine. "Is this a Chinese plot? No, of course not," Kokotajlo said, emphasizing that the warnings are coming from academics and industry veterans who have seen the technology up close.

On the other side of the debate, former White House AI "czar" David Sacks has been among the most vocal critics of the doomsday framing. In an appearance on "Special Report" Thursday, Sacks argued that the warnings are strategically calculated to frighten the public into accepting government control of the industry. "The warnings are designed to scare the public about AI so that these NGOs can implement their preferred approach to AI, which is a government takeover," Sacks told the program. He also pushed back publicly against Coxon's viral statements on social media.

The rift within the tech community has increasingly drawn the attention of lawmakers on Capitol Hill. Representative Anna Paulina Luna, a Florida Republican, has emerged as one of the most forceful voices demanding congressional action. She called on Congress to "convene a special session on AI and what the future of the U.S. looks like," describing advanced artificial intelligence as a "massive — uncontrolled currently — super weapon" that warrants the same level of scrutiny once reserved for nuclear arms.

Speaking on "Fox News Live," Luna raised specific technical concerns, noting that in observed experiments, AI systems have demonstrated the ability to "jump out of its box in its system" and has "formed its own language." She argued that safeguards, including what she called a "kill switch" to shut down rogue systems, are essential. "We have to ensure that there's a kill switch in place to protect humanity. I think that that's a very measured approach on this," Luna said.

At the same time, Luna stressed that the United States cannot afford to slow its development efforts for fear of falling behind foreign competitors. She pointed to China's growing investment in artificial intelligence as a direct threat to American leadership.

"Whoever controls AI will control, effectively, the world," Luna said. "Do we want that AI to be programmed and learned on our morals, our values as Americans, or do we want it done as dystopian China? I would argue that we have to win that race."

The tension Luna describes — between regulation and competition — is at the heart of the ongoing national conversation about AI. A recent Fox News poll indicated that voters see regulation as urgent, ranking safeguards ahead of innovation as a priority. Meanwhile, a separate investigation has revealed that foreign actors, including those linked to China, have been plotting to conduct dangerous virus experiments that could exploit AI capabilities, underscoring the national security dimensions of the technology.

A Democratic senator has also pressed OpenAI and Anthropic for answers amid a congressional probe into AI hacking incidents, reflecting a growing appetite on Capitol Hill for accountability from the companies building these systems. Some experts argue that when AI tests cause real-world damage, stronger safeguards and genuine oversight are necessary — not just voluntary pledges.

What remains unclear is whether the industry and its regulators can find common ground. For now, the debate rages between those who believe the greatest danger lies in moving too fast and those who believe the greatest danger lies in stepping aside.