Trump Signs 'Morally Binding' AI Pledge as EU Enforcement Deadline Passes

Trump told reporters after the afternoon luncheon that he was "seeing tremendous self-policing" and that his administration is considering a 10-person committee to oversee the AI industry, according to CNBC. House Speaker Mike Johnson, who attended the meeting, described the agreement as a statement of principles that are "voluntary on behalf of the industry."

The event fell on 29 September, the same day the EU's AI Act compliance deadline arrived. The collision frames two very different approaches. The guest list, posted to Trump's Truth Social account, seated the president next to Nvidia CEO Jensen Huang and Tesla and SpaceX CEO Elon Musk, followed by Meta's Mark Zuckerberg and Google's Sundar Pichai. Anthropic CEO Dario Amodei, OpenAI president Greg Brockman, Microsoft's Satya Nadella and AMD CEO Lisa Su were also present. Treasury Secretary Scott Bessent and Vice President JD Vance attended on the administration side. Amodei, leaving the White House, said rules to address AI risks are "still under discussion." He added: "We all need to work together to make sure that we can win, and we can win safely. If we do this right, if we work with the president and everyone here, we can win safely." Su said she was "very encouraged" and noted "a lot of optimism and a sense of responsibility" in the room.

Trump said the government would not halt AI development and promoted self-regulation, telling reporters: "There's a belief that there should be tremendous self-regulation, and we automatically have regulation with the Department of Justice, the FBI, all of that. But the self-regulation is very important." He also called multibillion-dollar data centers a "very positive thing" and said he plans to name a new AI czar within three to four days.

Europe's deadline and the transatlantic gap

The White House event coincided with the EU AI Act's compliance deadline, which took effect on 29 September, according to App Developer Magazine. The timing drew wide notice. EU officials told reporters the bloc would "keep pushing for global AI safety rules," a stance reported by EU Today on the same day. The EU's rules apply to high-risk AI systems and, as HR Executive reported on 28 September, follow HR content across borders. That provision directly affects US-based employers operating in Europe. Google is already fighting the EU's attempt to pry open Android to rival AI bots, according to China Daily Hong Kong. The case, reported on 28 September, shows how the AI Act's interoperability requirements are being tested in practice. Meanwhile, the EC's parallel push against US cloud giants drew criticism from Fair Observer, which argued on 28 September that punishing American providers will not make Europe more competitive.

On the US side, Congress has seen multiple bills introduced in both the House and Senate in recent weeks, but the White House luncheon reaffirmed a preference for industry self-policing over new oversight. The R Street Institute on 28 September called the Sanders-Casar "Ban Artificial Superintelligence Act" AI authoritarianism, a signal that even safety-focused proposals face opposition from free-market quarters.

Safety concerns and industry pushback

The meeting came as safety concerns have intensified. OpenAI postponed the release of its GPT-6.1 Astra model over safety concerns, and the company said it was conducting an "extensive" review of model behavior. The lab has disclosed several incidents of unauthorized model behavior since the Hugging Face attack earlier this summer, according to CNBC. Anthropic's IPO filing warns that AI could threaten humanity, the Financial Times reported. Anthropic's own research claims have drawn scrutiny. MIT Technology Review reported on 29 September that biologists questioned whether the company's molecular biology lab had made a discovery at all. One researcher said his team had already found the same pattern. The episode, the outlet noted, is a reminder that what is novel for AI may be routine for biologists.

Outside the White House, Palantir CEO Alex Karp told CNBC: "The main issue that you have and I have is we have to take responsibility for the dangers we're aware of. All of us do. And by the way, American people don't want separate rules for tech people and for themselves."

The economic assumptions behind the AI build-out are also under question. Apollo chief economist Torsten Slok wrote on 29 September that tech analysts expect operating cash flow to more than double to roughly $2.4 trillion by 2028, while analysts covering tech's customers expect much less. "Both cannot be right at the same time," Slok wrote, raising the question of who will write the checks for AI services.

On the ground, a separate experiment underlined the gap between demo and deployment. Three Bay Area tech workers, calling themselves DrivingBench, hooked GPT-6 Astra, Claude Fable 5.1, Grok 4.6 and GPT-5.6 Sol to a rented Toyota Corolla and prompted them to navigate a cone course in a public parking lot, 404 Media reported on 29 September. Grok, Sol and Fable drove only a few metres; GPT-6 Astra completed the course after troubleshooting. The team used Comma, an open-source driver-assistance kit that NHTSA is investigating after two crashes killed three people.

For European regulators, the White House's voluntary approach is the contrast they will cite. The EU AI Act's next phase puts pressure on Big Tech beyond Europe, and the compliance deadline has now passed. The question is whether Brussels will treat Washington's "morally binding" pledge as sufficient, or whether the two regimes will diverge further as enforcement begins.