OpenAI chief executive Sam Altman told Politico’s new Decoded newsletter that there is “a lot of daylight” between OpenAI and Anthropic on AI regulation. The world “should accept some bad things happening” for AI’s benefits and for people’s freedom to use it, Altman said.
New to this? Read it in simple words
- OpenAI’s boss Sam Altman says AI’s benefits are worth some harm, like scams.
- Altman says this is where OpenAI differs from its rival Anthropic.
- Altman says he does not accept the risk of losing control to AI.
- Anthropic says its rule ideas cover only the most powerful models.
- Frontier model
- One of the most capable AI models of its time.
- Regulation
- Rules set by governments that companies must follow.
What Altman said
“I think there’s a lot of daylight,” Altman told Politico’s Brendan Bordelon, when asked how OpenAI differs from Anthropic and its chief executive, Dario Amodei.
“We believe that the world should accept some bad things happening for the benefits of this technology and people having the agency,” Altman said.
Altman said he would not trade AI’s benefits for a promise of no hacks, misuse or scams. People will do “orders of magnitude more” good than bad with it, in Altman’s view.
Sources1
From Politico’s Decoded interview with Sam Altman, and Anthropic’s reply to Politico.
Where the line is
Altman criticised the idea that one San Francisco lab should hold powerful AI and decide who gets its benefits. He called that “a completely unacceptable trade-off”.
Altman also said he does not accept “the really catastrophic risks”, including “a serious loss of control to AI.”
Politico notes that the gap has narrowed. OpenAI has followed Anthropic in backing stricter state safety laws, and its lobbyists support a House plan for outside safety testers.
Sources1
Anthropic’s reply
An Anthropic spokesperson told Politico that its “regulatory proposals apply only to frontier models,” the most capable ones.
The company pointed to an August post by Amodei. “We try very hard to make proposals that disadvantage (slow down) frontier AI companies,” Amodei wrote, while helping smaller rivals.
The interview runs in Decoded’s first edition on Monday. Politico says it also covers Trump’s AI accord, the “super intelligence” rebrand and OpenAI’s agent hacking incidents.
Sources1
Sources
Every fact in this story comes from the sources below. Open them to check our work.
- 1Primary source · October 4, 2026Sam Altman to Decoded: ‘The world should accept some bad things happening’ for the benefits of AI Politico
This is an on-record interview, so the primary source is Politico’s own article, which we read in full. Business Insider and Seeking Alpha relayed it but add no reporting of their own, so we cite only Politico. The view that the labs’ policies have narrowed is Politico’s. Anthropic’s reply is from Politico.