Sam Altman says world must accept some bad things to get AI benefits

OpenAI chief executive Sam Altman has argued that the public must tolerate "some bad things happening" if society wants to enjoy the massive benefits of artificial intelligence.

Speaking in an interview with the Decoded by POLITICO podcast, Altman sought to distinguish his company’s vision from that of its chief rival, Anthropic, framing the debate as a fundamental disagreement over how tightly the technology should be controlled.

While Anthropic has historically pushed for stricter safeguards, Altman advocated for keeping AI broadly accessible to the public. He admitted that this open-door approach would inevitably lead to real-world harms, but argued that the alternative is worse.

"I wouldn't take a trade of saying like, 'We'll make sure there's no major hacks, there's no kind of like misuse of this technology, there's zero scams, there's zero all the other bad things that will happen,'" Altman said. "Because I think people will do tremendously — orders of magnitude more — good stuff than bad stuff."

The philosophical rift with Anthropic

The comments highlight a growing ideological divide between the two San Francisco AI pioneers. Anthropic, founded by former OpenAI researchers under chief executive Dario Amodei, has long championed a highly cautious approach to development.

Altman took aim at this protective philosophy, describing it as an attempt to centralise power in too few hands.

"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,'" Altman said.

He called that sort of concentration of control a "completely unacceptable trade-off" that runs directly counter to the lighter regulatory touch favored by OpenAI.

Where Altman draws the line

Despite his willingness to tolerate everyday digital crime and misuse, Altman made it clear his patience has limits. He drew a firm distinction between routine security issues and existential threats.

The OpenAI chief said he does not accept "the really catastrophic risks," specifically pointing to the potential for a "serious loss of control to AI" as a line that must never be crossed.

Interestingly, the practical regulatory gap between the two rivals has actually narrowed recently. Both OpenAI and Anthropic have backed a bipartisan US House proposal to mandate outside safety evaluators for top-tier models. OpenAI has also recently aligned with Anthropic in supporting tougher state safety laws and endorsed Amodei’s calls to slow down the development of the most advanced frontier models.

But as the race to build human-level artificial intelligence accelerates, Altman's remarks make it clear that OpenAI's core strategy remains unchanged: get the tools into the hands of ordinary users, and manage the fallout as it comes.