Trump calls the AI industry's own slowdown call a 'SICK conspiracy' -- his AI czar's answer is sharper: you don't need our permission, so just do it

President Trump used his Truth Social account on Monday to call the AI industry's own warning about itself a SICK conspiracy -- two days after Anthropic chief executive Dario Amodei proposed a deliberate slowdown in frontier AI capability, and one day after OpenAI's Sam Altman and xAI's Elon Musk both said in public that Amodei was right. “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, adding that his administration already has “tremendous CRIMINAL and REGULATORY power over these companies” and does not need new limits. He named Amodei directly, writing that the administration “has stopped AI 'people' from doing bad, or potentially bad, 'things,' like Dario (Anthropic!), who is now pretending to be a 'perfect little angel.'”

The Monday post followed a milder dismissal a day earlier, when Trump told reporters, “We're leading China in AI... and frankly, I want to keep it that way, because whoever wins AI wins,” waving off the specific dangers Amodei named as coming from “negative forces” pushing “things that won't happen.” Amodei's essay, published the Saturday before, argued close to the opposite: that AI agents given real autonomy are advancing fast enough -- through what he calls recursive self-improvement -- that a swarm of them could plausibly seize control of large parts of the internet within six to twelve months, causing what he estimated at hundreds of billions of dollars in damage without new safeguards.

Amodei's essay wasn't a one-line plea. It laid out a three-step plan, starting with outside evaluators getting employee-level access to frontier models -- a commitment Anthropic made immediately for itself, with Altman promising to match it "soon" and Musk offering three words of agreement and no commitment of his own. Microsoft separately published an actual governance document the same week, putting named prohibitions behind Satya Nadella's own endorsement of the pacing idea. Trump and Sacks were both responding to the headline ask -- slow down -- not to any of the specific mechanics underneath it.

The sharper answer didn't come from Trump. David Sacks -- the venture investor Trump appointed as the White House's AI and crypto czar -- posted his own response on X over the same weekend, and it reads nothing like the president's. “Dario has written that we need to 'pace the frontier,' and Sam has agreed... People may be surprised by my response: go ahead,” Sacks wrote, arguing that Amodei and Altman together hold close to a duopoly on frontier intelligence and therefore need no one's permission to slow down if they actually mean it. His pointed claim was about motive, not capability: stop pretending the motivation to slow down is purely altruistic, he wrote, arguing the real driver is the “massive product-liability exposure” either company would face if one of its models enabled a serious cyberattack.

The two responses land on the same bottom line -- no new federal restriction -- by two different roads. Trump's post treats the underlying risk as manufactured, attributed to unnamed “negative forces.” Sacks doesn't dispute that Amodei and Altman may be seeing something concerning in unreleased models; he disputes that seeing it requires a government process at all, since the companies could simply act on their own authority. (Sacks sits on the President's Council of Advisors on Science and Technology -- the same administration Trump was defending in the same 48 hours, making two different arguments from inside it.) That distinction matters for what happens next: an administration that thinks the risk is fake has no reason to act on it, and one that thinks the risk is real but belongs to Anthropic and OpenAI to solve has no reason to act on it either.

Neither response was built around this: the same week's reporting on Trump's post cited an NBC News Decision Desk poll finding that 70% of American adults say they worry more about AI than they feel excited by it, and 69% oppose a data center being built in their own community. On the numbers, public sentiment sits closer to Amodei's alarm than to either the president's dismissal or his AI czar's liability argument.

What the last 72 hours produced is three claims -- Amodei's, Trump's, and Sacks' -- doing very different amounts of work for how confidently each is stated. Sacks put his sharpest line under his own name, not anonymously:

“Stop pretending the motivation to slow down is purely altruistic.” — David Sacks, White House AI and crypto czar

Whether that description holds is worth checking claim by claim, since “purely altruistic” is doing a lot of work in a sentence that names no evidence for the alternative.

None of it settles Amodei's actual question: whether frontier labs are advancing fast enough to outrun their own ability to keep the result under control. Trump's answer is that the question is asked in bad faith. Sacks' answer is that the labs already have the authority to act on it and haven't used it. Both leave the burden exactly where Amodei's essay put it two days earlier -- on Anthropic, OpenAI and xAI to actually slow down, or to say specifically, not rhetorically, why they haven't.