Trump and AI executives sign voluntary controls document at White House

By Splitfeed EditorialLast updated Sep 29, 2026, 21:01 UTC

On September 29, 2026, President Trump hosted a White House lunch with a group of prominent AI executives including Meta's Mark Zuckerberg, Anthropic CEO Dario Amodei, and Nvidia's Jensen Huang. At the meeting, Trump and the executives signed a voluntary controls document described by at least one participant as morally binding and comparable to a constitution. Trump also announced an executive order that would rename artificial intelligence as superintelligence in all government documents, and he ruled out any joint US-China AI development venture, arguing such collaboration would disadvantage American companies.

The rename is the kind of symbolic move that tends to generate disproportionate reaction because it touches on how governments frame emerging technology for the public. Calling the technology superintelligence carries different connotations than artificial intelligence, implying a level of capability and urgency that remains scientifically contested. The voluntary nature of the controls document is equally significant: without enforcement mechanisms, its practical effect depends entirely on whether signatories choose to honor it. Geoffrey Hinton, widely referred to as the godfather of AI, publicly called Trump ill-advised on the subject, and White House economist Kevin Hassett argued AI creates rather than destroys jobs, a claim that remains disputed among economists and researchers.

The Pro-AI camp received the meeting positively, focusing mainly on Trump's refusal to partner with China on AI governance. The account Fllics quoted Trump's reasoning that such collaboration would make it harder for US companies to build better AI than their Chinese counterparts, framing competition as a reason to keep American AI development unconstrained. The Anti-AI camp was sharply critical, treating the event as a gathering of wealthy insiders shaping policy without accountability. Accounts including Manoj Upadhyay mocked the gathering as the world's most humble billionaires deciding human destiny, while others ridiculed the renaming as evidence of ignorance rather than strategy. The Middle Ground camp did not produce substantive reactions on the policy questions, though one account noted testing the government's existing AI chat tool for reliability, suggesting cautious evaluation rather than either enthusiasm or outright rejection.

The most concrete thing to watch is whether Trump signs the executive order formalizing the superintelligence rename and what, if any, language it includes about the voluntary controls framework. How agencies interpret and implement that order would be the first real test of whether the White House lunch produced anything beyond a symbolic agreement.