Sam Altman stresses regulatory worldview gap with Anthropic

OpenAI CEO Sam Altman stressed that a fundamental worldview difference with Anthropic on AI regulation remains. In an exclusive interview for the first edition of a tech newsletter and podcast launched on October 5 (local time) by a US political news outlet, Altman argued that some risks must be accepted for the benefits of AI and that the technology should remain broadly accessible to the public. The remarks came as the policy distance between the two companies has narrowed in recent months, and they stand out for how much effort Altman put into drawing a line amid the convergence.

Asked about his differences with Anthropic, Altman answered, "I think there's a lot of daylight." Asked specifically where he diverges from Anthropic CEO Dario Amodei, he said, "we believe that the world should accept some bad things happening for the benefits of this technology and people having the agency." He added, "I wouldn't take a trade of saying, 'We'll make sure there's no major hacks, there's no misuse of this technology, there's zero scams, there's zero all the other bad things that will happen,'" explaining, "Because I think people will do tremendously, orders of magnitude more, good stuff than bad stuff."

Altman also voiced the opposing argument himself. "I disagree, but I understand the perspective of people who are like, 'This technology is going to get so powerful, and it's so dangerous, that a single lab in San Francisco should have it and make sure nothing bad happens, and kind of figure out how to dole out the benefits,'" he said. He then called that approach "a completely unacceptable trade-off" that runs counter to the lighter-touch regulatory stance backed by OpenAI. According to reports, the remark drew on a regulatory capture criticism that tech advisers close to President Donald Trump have applied to Anthropic.

Still, Altman did not argue for accepting every risk. He later added that he does not accept the really catastrophic risks, including a serious loss of control to AI. According to reports, Altman said in an earlier, separate interview that a 10% risk of AI killing everyone by the end of the decade would be unacceptable, and that "I believe no lab has solved alignment."

The actual policy gap has been narrowing. According to reports, Anthropic has long been seen as the industry's strongest proponent of AI regulation, while OpenAI until recently favored a lighter approach. But OpenAI has moved toward backing tougher rules as it grapples with hacking incidents involving its models and rogue model behavior. Altman agreed last month with Amodei's call to slow the development of the most advanced AI models, and OpenAI followed Anthropic in endorsing new state laws with stricter safety requirements than before. OpenAI lobbyists also endorsed a bipartisan House proposal that would require top AI companies to embed outside safety evaluators. METAL has previously reported on Amodei's essay proposing a slowdown in AI development and embedded outside evaluators and Altman's statement pledging safety measures without an antitrust exemption.

Anthropic pushed back. An Anthropic spokesperson said the company's regulatory proposals apply only to frontier models and pointed to a social media exchange Amodei had with an investment manager in August. At the time, Amodei wrote, "Anthropic has always made its policy proposals very carefully," adding, "We try very hard to make proposals that disadvantage (slow down) frontier AI companies while advantaging smaller competitors."

The distance between the two CEOs showed beyond regulation as well. On October 3, Altman wrote on X: "I am very uncomfortable about people trying to ascribe religious force or a surrender of human judgment to AI models, and think it is a real safety issue." The post, which METAL reviewed, has passed 6.6 million views and 28,000 likes. According to reports, it was read as a veiled shot at Anthropic's work on the soul of AI.

The interview also covered Trump's new AI accord with major tech companies and the attempt to rebrand AI as superintelligence, autonomous hacking incidents carried out by OpenAI models and the company's legal liability, AI industry money flowing into US politics, and the influence of effective altruists inside OpenAI. Altman named French President Emmanuel Macron as the world leader who has impressed him most on AI policy, saying he "is extremely thoughtful and has been for a long time about where this is going and what the shape of it needs to look like." Macron cannot run in the April 2027 presidential election, and next month he will host the Paris Peace Forum with a focus on AI safety and the technology risks facing children and teenagers.

Through a tech law lens, the interview is a fight less over statutory text than over the philosophy of allocating responsibility. On concrete mechanisms such as slowdowns, outside evaluators and state safety bills, the two companies already stand on the same side. Where they split is how much residual risk, such as hacking, misuse and scams that those mechanisms cannot stop, society should bear, and whether strong regulation entrenches the position of a few labs. Altman drew a line between catastrophic risk and everyday harm, while Anthropic answered by narrowing its regulatory targets to frontier models. With legal liability for autonomous hacking incidents now on the interview agenda, where that line is drawn is becoming a question to be settled not in corporate statements but in legislatures and courts.