"We're prosecutors, not regulators, so we're not going to try to regulate AI. AI is extraordinarily beneficial to our society. There's a lot of great work that AI does, and it's also something that conveniently has become a flashpoint this week and the president has spoke about that and I very much agree with him."
On Monday, September 14, 2026, as recorded in entry #53, Trump posted to Truth Social that his administration "has stopped AI 'people' from doing bad, or potentially bad, 'things,' like Dario (Anthropic!)" and that "We already have tremendous CRIMINAL and REGULATORY power over these companies!" On Tuesday, September 15, Attorney General Todd Blanche briefed reporters in the Rose Garden and gave the answer above when asked about AI. He also said: "As far as AI is concerned, there's a lot of criminal laws. And so if anybody associated with AI violates criminal law, we'll investigate that. We already have." On Thursday, September 24, the White House held a state dinner for Chinese President Xi Jinping. Sam Altman of OpenAI, Elon Musk, Jensen Huang and Mark Zuckerberg attended and were photographed. Dario Amodei, chief executive of Anthropic — the man named in the September 14 post — was the one major AI chief executive absent, citing a scheduling conflict. Trump then personally invited him to a private dinner. That dinner took place at the White House at 10 p.m. on Sunday, September 27, the first one-on-one meeting between the two. Trump confirmed it. No readout was released; White House officials have not said what was discussed. Before the dinner Trump told reporters he and Amodei were going to "talk about" AI regulation: "I'm for let's go and let's win. You know we're about a year, maybe a year and a half up on China. There is nobody in third place. I mean it's just us and China." A White House official said only: "President Trump has been clear: America will lead the world in Super Intelligence, while protecting American consumers." On Tuesday, September 29, Trump and House Speaker Mike Johnson convene AI executives at the White House on AI safety and the China race. Amodei is not expected to attend; Anthropic co-founder Tom Brown is. Anthropic is reported to be preparing an initial public offering. The enforcement machinery referred to in the September 14 post is Executive Order 14365, signed December 11, 2025, which directed the Attorney General to establish an AI Litigation Task Force whose "sole responsibility shall be to challenge State AI laws."
Filed the dinner as a thaw. Axios had the scoop, CNBC confirmed it, the Washington Post framed it as the president dining with "the Anthropic CEO who called for AI slowdown," CBS noted there was no readout, and the trade press read it as access ahead of an IPO. All of that is accurate. Nobody put September 14 next to September 27. Thirteen days separate a written presidential claim of "tremendous CRIMINAL and REGULATORY power" over a named company and its named chief executive, and a private, unreadout, 10 p.m. meeting between the two men, arranged by personal invitation after that executive was the only one of his peers missing from the state dinner. Nobody asked what was said. And nobody set the Attorney General's answer against the post it followed by one day. "We're prosecutors, not regulators" was reported as reassurance, and it is the opposite of reassurance: the president had not claimed regulatory power alone, he had claimed criminal power, and the officer who holds that power answered by ruling out the lesser of the two. The sentence that went with it — "if anybody associated with AI violates criminal law, we'll investigate that. We already have" — was reported by one outlet and not followed up by any. Both men were asked about the same week, one day apart, and the word "criminal" survived in both answers without a single question attached to it.
This is filed as structural inference and not as documented, and the distinction is the whole entry. Every fact above is on the record: the post, the briefing, the absence from the state dinner, the personal invitation, the 10 p.m. dinner, the missing readout, the summit two days later, the IPO. What is not on the record, and may never be, is any connection between them. No charge has been filed against Anthropic or against Dario Amodei. No investigation of either has been disclosed. The dinner may have been exactly what the coverage says it was. The record simply establishes a sequence that would be reported as coercion in any other jurisdiction and was reported here as a scheduling story: a head of state writes that he holds criminal power over a named company whose chief executive has publicly disagreed with him, his Attorney General is asked about it the next day and does not disclaim the criminal half, and thirteen days later that chief executive is at a private table with him with no record of what was said. Entry #53 noted that the machinery in Executive Order 14365 was built to shield AI companies from state regulation and is now being invoked as leverage over them. This entry notes the next step, which is that leverage of that kind does not need to be exercised to work, and leaves no docket number when it isn't.