Two days after Dario Amodei asked the AI industry to slow down, the President of the United States called the whole idea a scam. In a Monday, September 14 post on Truth Social, Donald Trump wrote that “there is a SICK conspiracy going on against AI and Data Centers, and the only one that is happy about it is China,” dismissing the recent wave of AI risk warnings as a hoax and adding that “the only control or ‘guardrails’ that AI needs is a STRONG AND SMART (High IQ!) PRESIDENT, and the U.S.A. has that, in spades!” By the closing bell, the Nasdaq 100 had shed as much as 1.8%, and memory chip makers took the worst of it.

That’s the fourth straight week this site has tracked the industry’s pacing debate escalate, and it’s the first time it’s escalated into the Oval Office. What started as a researcher’s resignation letter, turned into a CEO’s public essay, turned into a delayed IPO, is now a fight between the sitting president and the three most powerful people in AI. Two of whom he appointed advisors from, or at least used to be friendly with.

Quick Summary: What Happened

Detail	Info
Monday, Sept 14	Trump posts on Truth Social calling AI slowdown warnings a “sick conspiracy”
His quote	”The only control or ‘guardrails’ that AI needs is a STRONG AND SMART (High IQ!) PRESIDENT, and the U.S.A. has that, in spades!”
Who he’s rebutting	Dario Amodei’s Sept 12 essay, endorsed by both Sam Altman and Elon Musk
Market reaction	Nasdaq 100 fell as much as 1.8% on the day
Hardest hit	Micron (-6%), SanDisk (-6%), SK Hynix (-7%)
Also pushing back	Former White House AI czar David Sacks, on separate grounds

Bottom line: The three loudest voices calling for AI to slow down just got publicly overruled by the one person whose guardrails they were arguing for, and the market didn’t like the fight.

What Actually Happened

Start with what Trump was actually responding to, because the timeline matters. On September 12, Anthropic CEO Dario Amodei published “We Must Pace the Frontier,” a three-step plan this site covered in detail arguing labs need embedded third-party evaluators, shared safety standards, and eventual government coordination on the most dangerous capabilities. Sam Altman backed it the same day — “I agree with Dario that we need to pace the frontier” — and Elon Musk offered two words: “Dario is right.” Two rival CEOs and the world’s most prominent AI-adjacent billionaire, agreeing on something publicly, in the same week.

That consensus lasted about 48 hours. Trump’s Monday post named Dario Amodei directly, mocking the Anthropic CEO for “pretending to be a ‘perfect little angel’” over his pacing essay, and calling the broader pushback against AI and data centers a “SICK conspiracy” that, in Trump’s telling, only benefits China. The fix, in his framing, isn’t industry pacing or outside evaluators — it’s him. “The only control or ‘guardrails’ that AI needs is a STRONG AND SMART (High IQ!) PRESIDENT, and the U.S.A. has that, in spades!” is a direct rejection of the entire premise of Amodei’s essay, which spends most of its length arguing that no single actor, president included, can substitute for independent verification of what frontier models can actually do.

His stated logic, per people familiar with the administration’s thinking, is competitive rather than technical: slowing down doesn’t make AI safer, it just hands the lead to China. That’s consistent with the deregulatory posture the White House staked out back in March, when it asked Congress to preempt state AI laws entirely rather than add new federal ones. Pacing, coordinated internationally or not, cuts against that framework by design — Amodei’s third step explicitly calls for the US to negotiate capability limits with “other states, including authoritarian ones,” which is a hard sell to an administration that’s built its AI identity around outrunning Beijing, not coordinating with it.

Trump wasn’t the only one pushing back. Former White House AI and crypto czar David Sacks argued the same day that Altman and Amodei don’t need new rules or a presidential guardrail to pace themselves — they already control the release valve on their own models and can slow down unilaterally any time they choose, without asking Washington for cover. It’s a different objection than Trump’s. Trump is dismissing the risk case outright; Sacks is accepting it in principle while arguing the fix doesn’t require government at all. Both land in the same place: no federal slowdown, no embedded evaluators mandated by anyone but the labs themselves.

Why Did AI Stocks Sell Off Over a Pacing Argument?

1. The pacing warnings themselves were already unsettling markets. Amodei’s essay and Altman’s IPO delay read, to investors, less like caution and more like the two best-informed people in the industry saying the technology might be moving faster than anyone can verify — that’s not typically bullish framing, even before politics entered the picture.
2. Trump’s rejection added political uncertainty on top of technical uncertainty. A president publicly calling his own AI leaders’ safety concerns a conspiracy doesn’t resolve the regulatory outlook — it just makes clear the White House and the frontier labs are no longer reading from the same script, and markets price uncertainty as risk.
3. Memory chips carry outsized exposure to AI infrastructure spending. Micron, SanDisk, and SK Hynix supply the high-bandwidth memory frontier labs need for training runs; any signal that the AI buildout could face turbulence — political or otherwise — hits those suppliers before it hits the labs themselves.
4. This is the fourth consecutive week of pacing-related headlines, following Jacob Coxon’s resignation, Amodei’s essay, and OpenAI’s IPO delay — each one alone might not move markets, but a monthlong drumbeat of “the people building this are worried” compounds.
5. A split inside Trump’s own orbit signals the administration isn’t unified on the response, which is its own source of investor unease — Sacks and Trump agree on the outcome (no mandated slowdown) but disagree on the reasoning, and that kind of internal incoherence tends to precede policy whiplash rather than settle it.

Why This Matters

Line this up against everything this site has covered since Coxon quit and the picture sharpens. Altman told OpenAI staff the company was open to pacing frontier development and asked Congress whether coordinating a slowdown with rivals would even be legal — a live antitrust question that gets more complicated, not less, once the president is on record calling the underlying safety case a conspiracy. If Trump’s position becomes the administration’s official line, the “ask Congress for a safety carve-out” strategy OpenAI was already pursuing has a much steeper hill to climb, because the White House isn’t just declining to help — it’s actively arguing the premise is fake.

It’s also the first time this pacing story has produced a same-day market reaction with actual numbers attached, rather than a stock analyst’s speculation about eventual impact. The IPO delay cost OpenAI a calendar year. This cost the sector a trading session. Different order of magnitude, same underlying signal: investors are starting to treat “the AI safety debate” as a variable that moves prices, not just headlines.

What Are Your Options Now

If you’re an enterprise buyer evaluating frontier AI vendors, nothing here changes what OpenAI or Anthropic ship to your contract this quarter. Trump’s post is a rhetorical rejection of the pacing framework, not a policy action — no executive order, no rule change, no funding decision has moved yet. Watch for whether this shows up in an actual White House statement or budget request before treating it as settled policy.

If you’re holding or trading AI-adjacent equities, memory and infrastructure names carry more political-headline risk right now than the model labs themselves, since they’re publicly traded and the labs largely aren’t. A 6-7% single-day move on a social media post, not an earnings report or a product launch, is a volatility signal worth tracking independently of your long-term thesis.

If you’re tracking AI policy for compliance or governance reasons, watch David Sacks’ position more closely than Trump’s. “Labs can self-regulate pace without government involvement” is the more durable policy argument of the two, and it’s the one more likely to actually shape whether Amodei’s second and third steps — industry standards, international coordination — ever get federal backing.

The Bigger Picture

The uncomfortable read is that this was never really a debate about whether AI needs pacing. It’s a debate about who gets to decide the pace, and Trump just answered that question for the administration: not outside evaluators, not international coordination, not the CEOs who build the technology — the president, unilaterally, framed as a guardrail nobody elected him to be in the way Amodei’s embedded-evaluator model was designed to be independent of any single person’s judgment, including a president’s.

That’s a genuinely different vision of AI governance than the one Amodei, Altman, and Musk endorsed on September 12, and the fact that it took the administration two days to say so out loud suggests it wasn’t a snap reaction. Calling a CEO’s safety essay a “sick conspiracy” isn’t the language of someone who disagrees on tactics. It’s the language of someone rejecting the entire category of concern — which puts the administration to the right of Sacks, who at least accepts the underlying risk case and just wants labs to self-police it.

Our Take

We think the market reaction tells you more than either side’s rhetoric does. Nobody sold off memory stocks because they suddenly discovered AI has risks — Amodei’s essay didn’t move markets two days earlier. What moved them was the president publicly fracturing with his own AI leadership on live television-adjacent social media, three weeks into a story about whether anyone can actually verify what these systems can do. That’s a governance-uncertainty trade, not a safety trade.

Sacks’ argument is worth taking seriously on its own terms, separate from Trump’s dismissal. “The labs can pace themselves without new rules” is functionally close to what Amodei’s own essay proposes for step one — Anthropic already committed unilaterally to embedded evaluators without waiting for government to mandate it. The disagreement isn’t really about whether self-regulation is possible. It’s about whether it’s sufficient without steps two and three, the ones that need other labs and other governments to show up. Trump’s post doesn’t answer that question. It just makes clear the federal government isn’t going to be the one asking it.

Frequently Asked Questions

What did Trump say about AI slowdown warnings?

In a September 14, 2026 Truth Social post, Trump said “there is a SICK conspiracy going on against AI and Data Centers, and the only one that is happy about it is China,” dismissing recent AI risk and pacing warnings as a hoax. He added that “the only control or ‘guardrails’ that AI needs is a STRONG AND SMART (High IQ!) PRESIDENT, and the U.S.A. has that, in spades!” — rejecting the case for independent oversight mechanisms like the ones Dario Amodei proposed two days earlier.

Who was Trump responding to?

He named Dario Amodei directly, mocking the Anthropic CEO as pretending to be a “perfect little angel” over Amodei’s September 12 essay “We Must Pace the Frontier,” which Sam Altman and Elon Musk both publicly endorsed the same day it published.

Why did AI stocks sell off after Trump’s post?

The Nasdaq 100 fell as much as 1.8% on September 14, with memory chip makers Micron, SanDisk, and SK Hynix among the hardest hit, down roughly 6%, 6%, and 7% respectively. The selloff reflects compounding uncertainty from a month of pacing-related headlines plus a new signal that the White House and frontier AI leadership are publicly at odds over how AI risk should be governed.

Did anyone else in the administration push back on Amodei’s pacing plan?

Yes. Former White House AI and crypto czar David Sacks argued separately that Sam Altman and Dario Amodei can control how fast their companies release capabilities without needing new government guardrails, since the pacing decision sits with the labs themselves. Unlike Trump, Sacks doesn’t reject the underlying safety concern — he argues it doesn’t require federal intervention to address.

Does Trump’s statement change any actual AI policy?

Not yet. It’s a rhetorical position, not an executive order, rule change, or funding decision. It’s broadly consistent with the deregulatory federal preemption framework the White House proposed in March 2026, which sought to override state AI laws rather than add new federal ones.

Is this the same pacing debate Anthropic and OpenAI have been having?

Yes. It’s a direct continuation of the industry-wide pacing conversation that began with a researcher’s resignation warning, continued through Sam Altman’s staff comments and OpenAI’s antitrust question to Congress, and led to OpenAI delaying its IPO to 2027. Trump’s post is the first time a government official has directly rejected the premise rather than debating how to implement it.

Does this affect enterprise AI deployments right now?

No. Nothing about Trump’s statement or the stock selloff changes product roadmaps, pricing, or contract terms for OpenAI, Anthropic, or other frontier vendors. It’s a governance and market-sentiment story, not a product story.

---

Last updated: September 15, 2026. Sources: Trump’s Truth Social post · Variety — Trump mocks Anthropic CEO, rejects AI guardrails · Dario Amodei — We Must Pace the Frontier · CoinDesk — Musk and OpenAI’s Altman agree with Amodei on AI slowdown · Axios — Washington’s AI paralysis: David Sacks on AI pacing.

Related reading: Dario Amodei Told AI Labs to Slow Down. OpenAI Blinked. · Altman Wants to Slow AI Down. Antitrust Says Maybe Not · Trump’s AI Policy: What Federal Preemption Means for You · Anthropic Caught Russia and China Weaponizing Claude