Sam Altman Says World Must Accept AI's 'Bad Things' as Leaders Sign Trump's Safety Pact

Sam Altman said the world must accept AI's 'bad things' as industry leaders signed a safety pact associated with President Trump, according to a Fortune report. The remarks and the signing landed together, tying a candid acknowledgment of downside risk to a formal industry commitment.

Altman's framing is notable for what it concedes. Rather than promising that AI will arrive without harm, he is quoted as saying society must accept the 'bad things' alongside the benefits — a stance that treats some level of negative consequence as unavoidable.

The pact itself — described in the report as Trump's safety pact — gives that public posture a collective counterpart, with industry leaders signing on rather than any single company setting its own safety line.

The combination matters because it captures a tension in current AI governance. Companies are simultaneously pushing to deploy more capable systems quickly and publicly committing to safety measures, and the two messages are not always easy to reconcile.

Given OpenAI's prominence, Altman's words carry weight beyond the company: they shape how other labs and policymakers talk about acceptable risk.

What to watch next is what the pact actually requires. A signed commitment is only as meaningful as its terms — whether it specifies testing, reporting, or disclosure obligations, and whether signatories are held to them.

The report frames the moment as an industry aligning behind a safety commitment while its leading figure publicly warns the public to expect some harm. The coming months will show whether the pact adds real accountability or remains a declaration of intent.