Altman says the world should accept ‘some bad things’ from AI as the White House accord draws scrutiny

Sam Altman said “the world should accept some bad things happening” in exchange for the benefits of artificial intelligence, in an exclusive interview with Decoded by POLITICO². He added that he disagreed with calls for much greater industry regulation³. The remarks were reported widely through Monday, with the OpenAI chief executive framing AI’s benefits as worth accepting some risks^4,6,7.

The timing ties the comments to policy. Altman said the world must accept AI’s “bad things” as industry leaders signed President Trump’s voluntary safety pact¹. That accord was signed by Trump alongside Google chief executive Sundar Pichai, Anthropic chief executive Dario Amodei, Meta chief executive Mark Zuckerberg and OpenAI president Greg Brockman⁸. Altman also said he sees significant daylight between OpenAI and Anthropic, a company that has built its reputation on pushing safety⁹.

Why it matters to investors

The industry is not speaking with one voice. The chief executive of Zscaler publicly rejected Altman’s framing, arguing the world should not accept some bad things from AI⁵. India is demanding binding AI rules, and the White House self-policing accord is drawing scrutiny for its voluntary character⁸.

A voluntary accord signed by the largest developers imposes no statutory compliance cost^1,8, while binding rules of the kind India is pressing for would⁸. Altman’s stated position puts OpenAI on the lighter-touch side of that line and away from Anthropic’s^3,9.

The accord’s signatories span the frontier labs and the platforms⁸, which is why Altman’s distancing from Anthropic on regulation is a commercial statement as well as a philosophical one^3,9.

What to watch

The first thing to watch is whether the voluntary accord holds as the sector’s main US guardrail, given that it was signed by Pichai, Amodei, Zuckerberg and Brockman at the White House rather than legislated^1,8. The second is jurisdictional divergence: India’s demand for binding rules is the clearest counterweight to the American self-policing model⁸.

The third is the split inside the industry. Altman has now put distance between OpenAI and Anthropic on regulation^3,9, and criticism from outside the frontier labs, such as the Zscaler chief executive’s response, shows the disagreement is not only between developers and governments⁵. Each position carries a different implied compliance bill for the companies investors own^2,7.