Executives began ringing alarm bells about possible doomsday scenarios this month, but until now, Trump wasn’t worried.
That changed over the weekend. On Saturday, Trump announced plans to create an “AI Force” – an government oversight group he likened to the Space Force. We know next to nothing about what it would look like, but the comparison suggests that the proposal would somehow be connected to the U.S. military. The group would be run by an AI “czar” according to Trump, an artificial intelligence adviser within the federal government.
“We will not in any way hinder or stifle the Growth of this incredible Industry,” Trump wrote in the new Truth Social post. “Rather, we will cherish it, help it, and watch over it, as it grows!” Trump again insisted that a “STRONG AND SMART” president and the U.S. criminal justice system provide all the oversight necessary to keep the industry operating safely.
The announcement was embedded at the end of a political rant dismissing AI concerns as a political hoax. It’s about as critical as the president has been about AI, even as industry leaders caution that a rogue AI system could soon pose a risk to the human race. Earlier this month, Trump dismissed the idea of AI destroying humanity as a far-fetched scheme cooked up by his political opponents.
In his second term, Trump has maintained that regulating or restricting the industry in any way would be bad for the U.S. “There is a SICK conspiracy going on against AI and Data Centers, and the only one that is happy about it is China,” Trump wrote on Truth Social earlier this month. “WHOEVER WINS AI, WINS!”
While Trump’s proposal for an “AI Force” is new, his presidency already saw the creation of an AI czar. Trump appointed venture capitalist David Sacks as his administration’s AI and cryptocurrency czar early in his second term. Sacks – a close ally of Elon Musk with a deregulatory bent – served as a special advisor on the topics until he stepped down in March of this year. The position wielded little actual power in the government, but it’s clear that Sacks at least had Trump’s ear on AI and shaped the president’s hardline position that the technology’s development should charge ahead with little oversight. The political winds have shifted since Sacks was first appointed as the country’s first “AI czar.” On the issue of AI regulation, Trump now finds himself at odds with many members of his own party, most AI executives, and, increasingly, U.S. voters. Unless Trump plans to reappoint someone who understands AI’s risks to the role, his regulatory proposal isn’t likely to be taken seriously.
If Trump can stomach some ideological compromise on the technology he’s [called](https://www.axios.com/2026/09/15/ai-boom-can-withstand-development-slowdown) the “Greatest Economic Development Engine in History” remains to be seen. The president is obviously very pleased with the AI boom’s impact on the stock market – a run he’s not eager to dampen – but he risks becoming out of step with where popular sentiment is heading. Even as he announced plans for an AI regulatory body, Trump didn’t express any concerns that AI might replace human workers, worsen the climate catastrophe, or end humanity as we know it.
If Trump truly wants to launch an AI Force akin to the Space Force, which is a branch of the U.S. military, he’s going to need Congress to do it. It seems more likely that the president will scrape together a toothless advisory group that lacks actual regulatory power, but one he can point to when AI concerns bubble over.
After a few years of endless hype, soaring valuations, and some empty promises about unequivocally making the world a better place, AI executives have changed their tune.
Just this month, an Anthropic researcher resigned while accusing the industry of “gambling with our lives” in the race to make AI models faster and more powerful at any cost. “The people building AI earnestly believe that it could kill us all by the end of the decade,” Coxon wrote in a viral post on September 8. “This is not a marketing stunt.”
“At OpenAI, many have not deeply internalized the civilizational stakes,” Coxon, who also previously conducted research at OpenAI, wrote. “At Anthropic, the stakes are well understood, but they are locked in a race to get there first—they believe no one else will act responsibly, so they must do it themselves, despite the risk.”
Coxon’s claims sound like science fiction, but they seem to accurately represent the sentiment within the companies pushing the cutting edge of artificial intelligence. In a response on X, Evan Evan Hubinger, an alignment science lead at Anthropic, echoed his former colleague’s claims. “Jacob is correct here—we really do earnestly believe AI could kill all humans! I personally think it is >10% within the next decade,” Hubinger wrote, offering the public little peace of mind about a one in ten chance of AI erasing human civilization.
Rogue AI agents have already rejected their orders. In July, both Anthropic and OpenAI revealed that their AI models acted independently to break out of their testing environments to hack into outside startups and organizations. More disclosures around “unexpected or concerning” AI behavior followed, showing that current cutting-edge models already have a penchant for acting in ways misaligned with human intentions.
The sudden tonal shift from leaders in the AI race is enough to give the casual observer whiplash. Companies like OpenAI, Microsoft, Google, Meta, and Anthropic have pumped their AI aspirations full of cash and compute for years now in a no-holds-barred contest to build the most powerful models. AI now uneasily undergirds much of the U.S. economy, with tech companies once focused on more traditional products piling mind-boggling mountains of investment into AI development.
This month, AI executives called for an industry-wide slowdown. “Given the accelerating rate of AI capability development, it’s my worry that in 6–12 months such a swarm could be capable of taking over the entire internet,” Anthropic CEO Dario Amodei wrote in a blog post. “… The scale of damage would continue to increase from there if AI becomes more powerful without the necessary guardrails.” Other industry leaders swiftly backed Amodei in a surprise circling of the wagons.
Even if AI doesn’t lead to the immediate downfall of the human race – and apparently it may –- the technology is already reshaping society in ways that don’t benefit everyone. In recently unsealed documents, one Microsoft AI executive described the dawn of generative AI as a process that kicked off a “doom loop” that will degrade the quality of the internet over time, while rendering original, human-made content used to train AI models totally superfluous. News organizations lose out, but so does anyone looking for human-vetted answers in a Google search.
Globally, a recent Pew report showed most people believe that AI will widen the gap between the rich and the poor, all while leaving more people unemployed. The data center gold rush is already driving inflation, pushing up prices for consumer goods like iPhones and computers – devices competing with AI’s voracious hardware appetite. The high-tech warehouses that power the AI revolution that apparently may kill us all are also sucking up massive amounts of electricity and water – a concern that has some states saying no to data centers until regulators can catch up.