A A new report from Politico makes it interesting, if somewhat amusing, to read now. The big AI wants the Trump Administration to speak with a clear voice about what’s wrong, and widely thinks it’s inevitable that it will eventually be hacked, as Anthropic recently did. But he also wants back the old Trump Administration — the one that said AI should not be regulated.
Dean Ball was recently hired at OpenAI in a position called “”.Head of Strategic Futures,” put it this way in Politico’s story:
“(T)here are things that the administration has done that I’m not a big fan of in terms of the acrimony, the opacity, the harshness, but more importantly, I’m glad they’ve decided to take this seriously.”
When I sat down to have my tooth pulled when I was about nine years old, before my dentist put an apron on me, he held up the instrument case. I remember a frighteningly huge, all-metal syringe; a pair of thick-walled pliers with textured handles like you’d see in a garage; and a surprisingly large, flat, blunt thing that looks like a tow truck driver can use it to open a locked car. These things were about to go into my mouth and my dentist was wise not to want me to be surprised.
OpenAI CEO Sam Altman’s visit to Congress in 2023 was in a somewhat similar vein, and he said“I think if this technology goes wrong, it can go horribly wrong. And we want to be vocal about that,” adding, “We want to work with the government to prevent that from happening.”
in his essay “Adolescence of TechnologyDario Amodei wrote that humanity’s ability to emerge from the chaos caused by artificial intelligence “will depend on our character and determination as a species, our spirit and our spirit.” And he added, “The coming years will be impossibly difficult, demanding more from us than we think we can give.”
this will hurt they seemed to say and if it does, we don’t want to be blamed. Altman and Amodei were not very sincere when they said these words, but they were also right.
The crucial difference between a great AI and a dentist is that America hasn’t actually asked Sam Altman or Dario Amodei to pull our metaphorical teeth. Dentists came unannounced. They don’t have certificates. And they make big promises about America’s smile that most of us can’t take seriously.
But you have to give it to them for one thing: at least they show us what’s on the plate.
And we don’t get involved. “Only 15% of Americans say they trust AI companies to make decisions about how AI is developed and used.” This is a quote Blog post on Anthropic about the request Conducted by Anthropic.
7 out of 10 people challenge data centers in our area. We are generally pessimistic about artificial intelligence and want the development to slow down.
Perhaps most importantly in today’s environment, 87% One of us says that within the next 20 years, “(f)oerign governments will use artificial intelligence technology to attack the United States.”
But when they took office, President Trump and Vice President Vance hinted at what you might call bravery about AI that the public doesn’t want. Its famous Paris speech in February 2025, Vance said essentially no regulation is coming and everyone better get used to it. Attempts to regulate artificial intelligence, he said, “will not only unfairly benefit those in space, but also paralyze one of the most promising technologies we’ve seen in generations.”
In other words, Vance didn’t care what was on the dentist’s tray. He had heard the promises of artificial intelligence dentists about America’s smile, and he didn’t want anyone but America to have the new, high-tech super smile.
For the most part, that’s been the Trump Administration’s stance on artificial intelligence ever since. A major brush with regulation earlier this month was when it declared Anthropic a supply chain risk. But it wasn’t that Anthropic’s metaphorical dental tools were scary and, in fact, potentially deadly. Because the Trump Administration loves them to be potentially deadly. He wants to say they’re the deadliest tools in the world—is that true or just a marketing gimmick—and he wants to be the only person with a say in how much and where lethality is met.
In other words, it was gratifying from a certain point of view finally Watch as the Trump administration fawns over Anthropic’s Claude Fable 5 model, one of the dentist’s tools, and put that whole unwanted dental procedure on hold for a moment.
“The administration’s current actions have put an almost complete moratorium on new releases,” Saif Khan, a former technology adviser to the Biden Administration, told Politico, adding, “And that’s going to start having a serious impact on companies’ bottom lines.”
Anthropic and its main competitor, OpenAI, are suffering together. As for the new family of AI models, the GPT 5.6 series is OpenAI sounding like everything is going mostly according to plan. That makes them available to a small group of VIP customers, and Anthropic will work with the Trump Administration to figure out how to roll them out without revoking privileges, as happened with the Fable 5 model earlier this month. But if you visit his blog postOpenAI is annoying. “We do not believe that this type of government access process should become a long-term default,” the company said.
“They’re walking on a bit of eggshells,” an anonymous policy adviser for frontier AI companies told Politico in a report.
But as my colleague pointed out yesterday:
“Meanwhile, cybersecurity experts were quick to point out after the Fable/Mythos ban that rival labs in China could develop their own AI to capture the breach, while labs in the U.S. try to figure out what’s allowed and what’s not.”
Earlier this month, more than a week before the export control directive on Anthropic models, the Trump Administration released an Executive Order requiring AI companies to submit their models to federal review. In a blog post about the GPT 5.6 models, OpenAI claims to be working with the Trump Administration to “develop a cyber Executive Order framework and a repeatable process for future model releases.”
But the US regulatory agenda has so far avoided the need for actual legislation by Congress to regulate AI’s capabilities and its role in our lives. Whether or not what’s good for AI companies is now a question of whether Donald Trump is happy with what he sees. He obviously doesn’t like guardians who can be arrested, like Tale 5 can and reportedly does. China-linked groups gain unwanted access to frontier models during periods when they should only be available to VIPs.
Dreams of what AI might do in future scenarios aside, we know what it does and we don’t like it. In other words, we’re now living in a painful future, the Big AI CEOs warned us about, and the president finally stopped the dental procedure. The problem is, it’s about to change nothing. and restart.
