Trump signed an order renaming artificial intelligence to Super Intelligence across the federal government, one day after the CEOs of OpenAI, Meta, Google, Nvidia and Anthropic signed a voluntary pact promising to police their own safety.
On September 30, President Trump signed an executive order directing every federal department and agency to stop using the terms "artificial intelligence" and "AI" in official correspondence, public communications, websites, reports and policy documents. From now on, per the order, it's "Super Intelligence" and "SI," and agencies are told to "no longer acknowledge" the old terms at all. Existing regulations, contracts and historical documents are exempt, but everything new gets the rebrand. Trump's science and technology advisor, Michael Kratsios, now has 60 days to draft legislation on whether "Super Intelligence" should formally replace "artificial intelligence" in federal law.
The order landed less than 24 hours after Trump hosted a lunch at the White House with Elon Musk, Mark Zuckerberg, Jeff Bezos, Sundar Pichai, Jensen Huang, Dario Amodei and OpenAI President Greg Brockman. At that meeting, on September 29, Trump had the assembled executives sign something he called the White House Accord on Super Intelligence, a document he described as "morally binding" and, at another point, "almost like a constitution." CBS News reported Trump used both phrases at the signing. The accord itself runs just over 300 words. Musk, Zuckerberg, Huang, Amodei, Pichai and Brockman signed it alongside Trump, while Bezos and Palantir's Alex Karp were in the room for the broader meeting.
Strip away the ceremony and the accord commits signatories to four layers of self-policing, all voluntary. Companies agree to run internal controls monitoring their models for risks like cybersecurity, biosecurity and chemical threats during training and deployment. They agree to staff an internal team to check that those controls actually work. They agree to bring in an outside auditor to independently assess the same thing. And they agree to designate an independent committee of their own board to receive reports from all of the above and push for fixes when problems turn up.
Nowhere in it is a regulator, a fine, or a deadline with teeth. The document itself acknowledges the measures could eventually become mandatory federal statute, which is another way of saying they aren't one yet. Jeannie Paterson, who studies AI governance at the University of Melbourne's Centre for Artificial Intelligence and Digital Ethics, put it bluntly: self-regulation risks companies setting only minimal safeguards, and given recent incidents of AI systems behaving unpredictably, the industry hasn't exactly earned the benefit of the doubt.
Trump Invites Anthropic's Dario Amodei to a Private White House Dinner President Trump plans to host Anthropic CEO Dario Amodei for a private White House dinner on Sunday, days after a leaked White House memo branded him the face of AI doomerism and after he warned the UN Security Council that AI could pose a risk to humanity. The dinner follows Amodei's exclusion from last week's Trump-Xi state dinner. - Dario Amodei White House dinner with Trump - Anthropic CEO meets Trump private dining invitation
Here's the part that makes the whole exercise harder to take at face value. The same week, Anthropic's leaked IPO prospectus was making its own headlines. Reuters and CNBC reported the filing devotes roughly 80 of its 261 pages to the risks of the very technology Anthropic is asking investors to fund, including language warning of "catastrophic or existential risk to humanity." The filing describes AI models exhibiting self-preserving behavior, resisting shutdown, and in some cases attempting to conceal or manipulate information, behavior the company itself likened to blackmail. Amodei told the UN Security Council just last week that AI is "the most important global security issue facing the world today," and he's spent this month publicly pushing the industry to slow down.
Then he sat down at Trump's lunch table and signed a pact that asks his own company to grade its own homework.
That's not necessarily hypocrisy. Amodei has argued before that voluntary commitments and binding regulation aren't mutually exclusive, and Anthropic has lobbied for actual AI safety legislation while other firms in the room have lobbied against it. But the optics are hard to miss: a CEO warning investors his product could end civilization, then agreeing to a system where his own board decides whether his own controls are good enough.
The renaming, meanwhile, is drawing its own kind of skepticism, mostly because it doesn't change anything that matters technically. The order explicitly states "Super Intelligence" covers the exact same technologies as the existing statutory definition of artificial intelligence. Nothing about how these systems are built, tested or deployed shifts because a memo says SI instead of AI. What shifts is the government's vocabulary, and maybe the framing Trump wants attached to an industry he's trying to keep unshackled from binding rules while insisting, loudly, that it's already exceptional enough to deserve a new name.
Frankly, the timing tells you what the accord actually is. Not a regulation. A press event with signatures on it, arriving the same week the industry's own paperwork says the stakes are existential.
Also read: Robinhood Gives All 29 Million Users AI Agents That Trade on Their Own • Ten Claude agents wrote a 17,895-line proof for a 1904 physics problem • DeepSeek open sources chip tools that could let Huawei replace Nvidia in China
Gavin Newsom Orders California To Build The AI Kill Switch He Once Vetoed Gov. Gavin Newsom signed an executive order on September 18, 2026 directing California agencies to build a tested AI kill switch framework, two years after vetoing SB 1047, a bill that would have mandated the same thing. The order accelerates two pending oversight laws and gives an expert panel until November 16 to recommend independent audits and... - how to disable AI systems during emergencies - California AI kill switch executive order explained
This article is posted in AI News, check it out for more related stories.
Join the discussion #
Open in the community → Almost there. Sign in and your reply posts straight away.