In an interview for the debut edition of Decoded by POLITICO, published Sunday evening, Sam Altman spelled out where he thinks OpenAI and Anthropic still part ways. Asked specifically where he diverges from Anthropic and 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.” A moment earlier he had sized up the gap himself: “I think there’s a lot of daylight.”
Where Altman draws the line
To make the point, Altman reached for a charge that tech advisers around Trump level at Anthropic: regulatory capture. His version of it: “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 called that “a completely unacceptable trade-off.”
Then came the line everyone is quoting: “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.’ Because I think people will do tremendously — orders of magnitude more — good stuff than bad stuff.”
He did draw a limit of his own. The “really catastrophic risks” are off the table, including “a serious loss of control to AI.”
Anthropic replies in a single sentence
POLITICO asked Anthropic to respond. A spokesperson said the company’s regulatory proposals apply “only to frontier models.” Beyond that, Anthropic pointed to an exchange Amodei had with an investment manager back in August: “Anthropic has always made its policy proposals very carefully. We try very hard to make proposals that disadvantage (slow down) frontier AI companies while advantaging smaller competitors.”
That was it. No pushback on Altman’s risk math, just a note about scope.
The positions are converging, the rhetoric isn’t
What makes the interview interesting is the context POLITICO puts next to the quote. Altman agrees with Amodei’s September call to slow development of the most capable models. OpenAI has backed state laws with stricter safety requirements than anything it supported before. And OpenAI lobbyists have endorsed a bipartisan House proposal that would make large AI companies bring in outside safety evaluators.
So the substance has moved closer together while Altman works to keep the distance visible. That tracks with what we saw in late September: Trump held to his no on a statutory brake over dinner with Amodei, and days later Amodei’s name was on the voluntary accord at the White House anyway. To be heard in Washington right now, you apparently have to do both — agree, and set yourself apart.
The interview covered more ground: Trump’s rebrand of AI as “super intelligence,” the run of autonomous hacks by OpenAI models and the liability question behind them, industry money in US politics, and the influence of effective altruists inside OpenAI. It ran in Monday’s edition of Decoded and the accompanying podcast.
Sources: