Trump Wants AI Firms to Police Themselves
At a Washington meeting the US president backed industry self-commitment over outside oversight; the agreed principles were not spelled out.
In short
Trump's position is that oversight of AI belongs with the companies themselves, because in his account their own commercial exposure already forces them to prevent harm.
At a glance
- Tuesday meeting in Washington: Trump and the heads of the AI industry, with regulation as the main topic.
- Trump's line: the companies supervise themselves rather than answer to an outside body.
- His reasoning: the firms' own economic stake is protection enough.
- Outcome: a set of self-commitment principles whose wording the available report does not give.
- Unverified: the exact date, the attendance list, the text of the principles — two of three sources would not load.
Trump's answer to AI oversight is to leave it where the systems are built. Meeting industry leaders in Washington on Tuesday, the US president argued that AI companies should police themselves instead of answering to an outside authority. The meeting produced a set of principles the participants commit to voluntarily, according to the report in Austria's Der Standard.
What the meeting actually produced
Not binding rules. A self-commitment holds for as long as the signatories want it to hold — there is no body checking, and no consequence attached to walking away. The report we could open does not list what the principles cover.
That leaves the question that matters in practice unanswered. Whether any testing or disclosure duty follows from them is open, and we will add the detail once the text is public.
The argument: commercial exposure as the safeguard
The president's case is economic rather than legal. On his account, anyone who has built a company of this size will not gamble it on a product that causes harm.
"Your companies are at stake if something goes wrong," Trump said, in the German wording reported by Der Standard.
He added, in substance, that the firms would therefore stop harm on their own. The effect is to move the judgment call from lawmakers into the vendors' own risk calculus.
The objection: existential risk
The report sets Trump's stance explicitly against warnings about existential AI risk. Part of that warning has come from the industry now being handed responsibility for its own supervision, including vendors that market safety research as part of their identity.
That is the tension in the room. The same companies that describe their technology as dangerous would get to draw the line themselves.
What we could not verify
Of the three sources planned for this piece, only one could be retrieved, and only as far as its paywall. Two of three sources would not load — including the reporting that the guest list comes from.
Earlier coverage framed the event as an AI summit at the White House, with the heads of Google, OpenAI and Anthropic among those invited. Which of them attended is not confirmed in the source available to us, and neither the exact date nor the wording of the self-commitment is documented. We are not treating any of that as settled.
FAQ
What did Trump decide at the Washington AI meeting?
No legislation and no new regulator — a self-commitment instead. Under Trump's approach the AI companies supervise themselves, and a set of principles was agreed at the meeting. The wording of those principles is not given in the report available to us.
Does self-regulation mean there are no AI rules in the US?
That does not follow from the report. It describes Trump's preference for how oversight should work going forward, not the repeal of existing requirements. A voluntary commitment sits alongside the law rather than replacing it.
Which AI companies were at the meeting with Trump?
Earlier reporting said the chief executives of Google, OpenAI and Anthropic were among those invited. The source we were able to open names no attendees, so the presence of any individual is not confirmed.
