· via TechCrunch
White House renames AI 'super intelligence' in executive order with non-binding safety pact
President Trump ordered U.S. officials to call AI 'super intelligence' while top AI CEOs signed a voluntary safety pact, an event TechCrunch's podcast team reads largely as a rebranding exercise for the industry.

An executive order on vocabulary
The U.S. government has a new official term for artificial intelligence. According to TechCrunch, President Donald Trump signed an executive order this week directing official U.S. representatives to say "super intelligence" instead of "artificial intelligence," and he announced the change at a gathering of leading AI executives.
The rebranding did not stop there. TechCrunch reports that Trump followed up over the weekend by announcing a new "Super Intelligence Force." The event itself brought together an unusually concentrated roster of tech power brokers: Mark Zuckerberg, Jeff Bezos, Elon Musk and Anthropic chief executive Dario Amodei were all in attendance, and the joint press conference afterward drew its own wave of online commentary.
A voluntary commitment
The formal centerpiece of the gathering was the "Joint Commitment on Frontier Responsibilities," signed by the executives present. On TechCrunch's Equity podcast, the hosts were broadly skeptical of its substance. Sean O'Kane described the document as "deeply non-binding" from a legal standpoint and entirely voluntary, while Trump characterized it as "morally binding" — a framing O'Kane said he had never encountered before.
What made the ceremony notable was how quickly it came together. As TechCrunch's Anthony Ha observed, the president had been saying only weeks earlier that the United States should lead in AI without constraining it and that anyone warning of danger was essentially arguing in bad faith. The pivot to a safety-themed signing event suggests the administration saw value in the gesture, even if the agreement itself is unlikely to change how companies behave. The apparent haste showed in the details: O'Kane noted that the signed document misspelled "United States."
Anthropic's partial rehabilitation
The most striking subplot involved Anthropic. TechCrunch reports that the company had been out of favor with the administration until shortly before the event, a shift that began when Amodei was invited to a 10 p.m. dinner with Trump. Days later he was seated alongside the other executives at the luncheon.
The repair may be incomplete. O'Kane pointed out that the Defense Department continues to describe Anthropic as a national security risk, and that Emil Michael has been repeatedly attacking the company publicly on X. Ha added that earlier reporting on the dispute suggested a divided White House: the Treasury Secretary was described as sympathetic to Anthropic, while defense officials objected to conditions the company wanted to attach to the use of its technology. On that reading, little has fundamentally changed — parts of the administration want Anthropic involved in these decisions, and others firmly do not.
The rebranding is the story
For Kirsten Korosec, the meeting's lasting significance lies in the messaging rather than the paperwork. She characterized the event as a rebranding exercise built around a clear separation: the old label carries the familiar anxieties about existential risk and job losses, while "super intelligence" is positioned as something apart from those fears. Whether the public draws that distinction is another matter.
Why it matters
A government-mandated vocabulary change is unusual in itself, and it signals how the administration intends to frame its AI agenda — leadership and ambition under a fresh name, rather than risk management under a loaded one. For AI companies, the commitment creates no new enforceable obligations, yet the gathering shows that even a regulation-averse White House saw political value in being seen to act on AI safety, which the TechCrunch panel read as a response to the technology's image problem. The mixed signals around Anthropic are a reminder that the administration is not a single actor on AI policy, and that one dinner does not settle disputes between agencies. Anyone tracking U.S. AI policy should watch both the language officials are told to use and which parts of the government are doing the talking.
- #ai-policy
- #white-house
- #ai-safety
- #regulation
- #artificial-intelligence