The Cowardly Capitulation of AI’s Top Leaders to Trump’s Absurd “Super Intelligence” Rebrand

Seventy summers ago, a small group of pioneering scientists gathered at Dartmouth College to debate a radical, almost unthinkable premise: that one day, machines could develop thought processes indistinguishable from human cognition. One attendee, John McCarthy, proposed a name for this fledgling, untested field: artificial intelligence. The label stuck, for 70 full years.

Type “artificial intelligence” into Google Scholar today, and you’ll pull up roughly 7.42 million published results. The term is completely ubiquitous. Everyone recognizes it, even people who deeply resent AI and its societal impacts — for most critics, arguing over what to call the technology is the absolute least of their concerns.

Well, almost everyone. For some inexplicable reason, President Donald J. Trump developed a personal grudge against the decades-old term. When he dislikes something, he has never hesitated to throw out arbitrary edicts to erase it. During a September 23 address at the United Nations, he dropped a clear hint that he wanted to scrap the “artificial intelligence” label entirely. At the time, his offhand rant read like typical bluster, even for a president famous for arbitrarily renaming gulfs, lakes, and public arts centers. Rechristening an entire global scientific field and America’s most strategically critical economic sector? That seemed far too ridiculous to ever actually happen.

But this week, during a closed-door White House lunch with dozens of the most powerful leaders in the U.S. AI industry, Trump followed through, signing an executive order that declared “artificial intelligence” a verboten term for all federal government use. From that point on, all official government communications would refer to the technology exclusively as “Super Intelligence.” The president made no secret that he expected the entire private sector to fall in line behind the rebrand.

In a display that will go down as an embarrassing mark on the entire industry, AI’s most powerful titans meekly went along with the demand. For all intents and purposes, the so-called lords of AI have now become the lords of SI.

It is impossible to miss what this moment actually is: the ultimate presidential power play. No matter how many trillions of dollars these CEOs control, no matter how much genuine technical intelligence they have built their careers on, the undisputed alpha in that lunch room was the famously orange-skinned commander in chief.

Changing a few words in a name might seem like a trivial, meaningless detail next to the far bigger, existential challenge of regulating a technology that carries clear catastrophic risks if misaligned. But for this room full of AI leaders, “artificial intelligence” was never just a random string of letters. It was the foundation of their field, the core of their professional identity. It was the phrase that sparked their imaginations when they stayed up late reading science fiction novels under their covers as kids. They built and remade their multi-billion dollar companies explicitly around advancing AI, and the term has dominated every public discussion they have had for the last four years straight.

Yet when Trump used that lavish lunch gathering to strong-arm them into endorsing his absurd plan to undo 70 years of established scientific history, not a single person in the room spoke up in objection. Not Jeff Bezos, not Jensen Huang, not Dario Amodei, not Mark Zuckerberg, not Sundar Pichai, not Greg Brockman, not Alex Karp, not Elon Musk.

Jacob Weisberg, author of the recent book Profiles in Cowardice: A Study of Collaboration in the Trump Era, frames the entire episode as a classic Trump loyalty test. The point of the demand, Weisberg argues, was never about the name itself: it was a chance to see if even the most powerful, independently wealthy executives in the country would willingly debase themselves to curry presidential favor and avoid drawing his punitive wrath. Pass Trump’s loyalty test, the old adage goes, and you automatically fail the basic backbone test.

“The only way to push back against this kind of arbitrary, petty demand is collective action,” Weisberg points out. “If every single one of those leaders had simply said, ‘We’re happy to agree to your proposed self-regulation terms, but we can’t accept this name change,’ what on earth would he have done to them?”

Instead, the billionaires’ sheepish, total silence let Trump claim publicly that the world’s top AI experts had enthusiastically endorsed the rebrand. When a reporter in the press gaggle after the lunch referenced AI agents, he immediately cut her off to correct her. “The name has been changed by the biggest, the smartest, the greatest people anywhere in the world,” he insisted.

That claim is entirely false. Even using the full capabilities of so-called superintelligent systems like ChatGPT or Gemini, you cannot find a single public statement from any of those CEOs arguing that the term “artificial intelligence” was outdated and needed to be scrapped. Satya Nadella once offhandedly noted he would have preferred the field be called “different intelligence,” but he has never stopped using the standard AI term in all of Microsoft’s public and product communications. Mark Zuckerberg recently launched an internal Superintelligence Lab, but Meta still markets a flagship consumer product called Meta AI to billions of users, and sells a line of consumer hardware explicitly branded as AI glasses.

I reached out to Google, Microsoft, Anthropic, Meta, Amazon, OpenAI, and xAI to ask if they planned to scrub all references to “AI” from their product names, research papers, and public communications. I even asked OpenAI and xAI if they intended to rebrand their entire companies to comply with the presidential order. As of this writing, not a single one of them has responded to my query. My best guess is that they will do the absolute bare minimum to avoid drawing Trump’s anger, while dragging their feet until the next presidential administration takes office and throws this bizarre, unworkable mandate in the trash.

Some observers have argued that the executives might actually welcome the rebrand, since “artificial intelligence” has increasingly become a loaded, toxic term in public discourse. I remain deeply skeptical of that take, especially since Trump’s proposed replacement, “super intelligence,” is objectively far scarier to the general public than the original, long-familiar label.

From a purely scientific perspective, the rebrand makes no logical sense at all. No less an authority than Nick Bostrom, the philosopher who popularized the term “superintelligence” with his landmark eponymous 2014 book, told me he is completely baffled by the mandate. “Superintelligence refers specifically to hypothetical systems that far outperform the very best human brains across every possible cognitive task,” Bostrom explained over email. “The term ‘AI’ covers a vastly broader, much more established category of systems: that includes spam filters, spell checkers, basic navigation tools, early large language models, and even the chess engines built back in the 1970s that did not yet beat the top human grandmasters. It would be deeply awkward to call every single one of those things ‘superintelligent.’”

So besides the fact that he could, what was Trump’s actual stated rationale for demanding this massive terminology shift? In his UN speech, he claimed he hated the word “artificial” because it implied the intelligence was fake — and he insisted the technology’s capabilities are very real. After the White House lunch, he doubled down, declaring that the term “artificial intelligence” itself was the fake thing: “It’s a fake word, it’s no good, it’s an inappropriate term.”

You don’t need a degree in psychology to spot the clear throughline between this impulse and his recent decision to bar CNN, Politico, and MSNow from White House press access, which he justified by claiming their news coverage was “fake.”

That said, the logic that equates the word “artificial” with “fake” doesn’t hold up to even basic scrutiny. When John McCarthy coined the term back in 1956, he used the word “artificial” specifically to distinguish machine-generated intelligence from the natural, biological intelligence that human beings possess. The biggest public risk right now is not that people think AI outputs are fake — it’s that the outputs are so convincing, most people can’t tell if a given piece of text, image, or video was made by a human or an AI. Mandating clear, explicit distinction between the two with required watermarks and labeling should be a core part of responsible AI regulation — the exact kind of oversight the Trump administration has consistently refused to implement.

As Weisberg notes, this kind of absurd, unprompted arbitrary edict is completely par for the course for Trump. What is far more disturbing is the total, glaring lack of character from every single top AI executive in that room who refused to push back. They walked away from the lunch with the big prize they have been lobbying for for years: voluntary industry self-regulation, in place of strict federal oversight. But they left their dignity at the White House door.

Worse, their total failure to defend the basic right to use the 70-year-old standard name for their field — something California Governor Gavin Newsom pointedly did in a trollish counter executive order reaffirming the state’s use of the term “artificial intelligence” — throws the entire self-regulation deal into serious doubt. Making voluntary safety standards actually work requires enormous fortitude, because prioritizing safety almost always runs directly counter to short-term financial and competitive incentives. Yet the public is now being asked to trust these supposedly independent, industry-leading corporate giants, who all nodded along like bobblehead dolls while a technologically illiterate president demanded they abandon the term that has been the core of their careers, their companies, and their entire livelihoods for decades.

That is not just inappropriate. It is fake.

This is an edition of Steven Levy and Lauren Goode’s Backchannel newsletter. Read previous issues of the publication here.