President Donald Trump met with top U.S. technology executives on Tuesday and made his priority plain: keep American AI moving. “This is one more example of why I will never stifle the growth of technology that will be bigger than the Industrial Revolution,” Trump said. The meeting came as concerns grew that AI models could lose control and researchers and lawmakers called for stronger safeguards.
The White House session followed Trump’s unveiling of a new A.I.-powered website designed to help Americans navigate government services. He’s also pushing a new name for the technology: “superintelligence,” according to PBS News Hour host Geoff Bennett.
The China Calculation
Trump framed the stakes as a matter of national advantage. “President Xi knows that. We’re leading by a lot,” he said. He rejected the idea of a joint venture with Beijing, which he said some people had floated. “If we do that, we’re going to be giving a lot of secrets that -- that we have that they just would want to have or anybody else,” Trump said. “And, frankly, there is nobody in third place. I mean, it’s all between us and China.”
Bennett said Trump appeared focused on industry self-regulation at the meeting. Technology companies signed a voluntary agreement focused on AI safety reviews.
The Case for Outside Rules
Gary Marcus, professor emeritus at New York University and author of “Taming Silicon Valley,” told Bennett on the PBS News Hour that the voluntary approach falls short. “It’s certainly not enough. And trusting these companies is insane,” he said. He cited a New York Times report that OpenAI knew about problems behind the Hugging Face incident months before it happened but did nothing because it wanted to rush products to market. Marcus wants outside regulation, as in aviation, food and drugs.
His argument had limits, though. Marcus said he doesn’t think extinction is “a real likely risk for humanity,” and called Terminator-style scenarios fantasies. The danger, he said, is more ordinary: catastrophe caused by errors, hallucinations and cyber hacking. He cited an incident in which an AI reported something incorrectly and nearly led to war. He also described a hack that could temporarily take down an electric grid and then a hospital. Most incidents wouldn’t be harmful, he added, though one could be.
Marcus also called the race with China “a little bit overblown.” China, he noted, has regulation and a licensing regime for its software products. He predicted the contest would end in a tie, like Coke and Pepsi, with each country holding its own “sovereign AI.”
A Secret Framework
The sharpest detail in Marcus’s account concerned how Washington already acts. He said the White House does some pre-release vetting without independent oversight, using a secret framework it hasn’t revealed to the public. According to Marcus, the White House precluded and delayed Anthropic’s Mythos system but hasn’t delayed OpenAI’s latest system, which he thought had more problems in some ways.
Marcus called the process “idiosyncratic” and not transparent. Customers, he said, can’t tell what will be released and can’t trust U.S. AI. His remedy: a dedicated federal AI agency that would independently evaluate new products, modeled on drug approval. Makers would have to show that benefits outweigh risks. He also wants audits after release and full disclosure of incidents.
Asked whether Washington could act before a major crisis, Marcus called it “the, you know, what is it, $3 trillion question.” The Hugging Face hack, he said, along with tens of thousands of similar incidents, was a wake-up call.
Why This Matters: The debate pits speed against structure, and both sides’ own words show the tradeoffs. Trump sees a two-horse race with China and won’t risk the lead or hand over secrets. Marcus, an industry critic, concedes that extinction is unlikely and most incidents aren’t harmful. He also says the current arrangement is a secret, discretionary process that leaves customers guessing and treats two companies differently. From a market-minded view, unclear rules leave businesses and buyers to plan around uncertainty. A new federal agency would add its own price tag, while the drug-approval model Marcus cites raises questions about delay. Whether Congress chooses public, predictable criteria or leaves the White House’s unpublished framework in place will shape how fast American firms can build and sell.