Trump Brands the AI Slowdown Push a 'Sick Conspiracy'

President Donald Trump called demands for a slowdown in artificial-intelligence development a "sick conspiracy," joining investor Michael Burry and others in criticizing Anthropic's chief executive, who recently argued that AI companies need stronger safety measures.

The argument over how fast artificial intelligence should be built has stopped being a technical debate among researchers and become a political one. President Donald Trump described calls for an AI slowdown as a "sick conspiracy," and investor Michael Burry is among the others publicly criticizing Anthropic's chief executive, who recently said AI companies need to adopt better safety measures.

That is an unusual alignment. Trump's objection is framed in terms of national competitiveness and the risk that American developers cede ground by voluntarily easing off. Burry's is the objection of a short-seller-turned-skeptic who has spent the past year questioning the economics behind the AI trade. They arrive at the same target from opposite directions, and the target is the most safety-forward voice among the frontier labs.

Why a safety argument became a market argument

Frontier AI labs sit at the center of an enormous capital-spending chain. Data-center construction, power procurement, advanced packaging, networking and memory all price off an assumption that model development keeps accelerating and that compute demand rises with it. A credible call from inside the industry to slow the frontier down is therefore not just an ethics debate; it is a challenge to the demand curve that underwrites a large slice of equity-market capitalization.

That is what makes the pushback so vigorous. If the leading labs were to adopt a genuinely slower release cadence, with longer evaluation windows and outside testing before deployment, the near-term consequence would be a stretching of timelines. Revenue that was modeled to arrive in one year arrives in two. Capital already committed to buildings and silicon sits idle for longer. Nothing about that is fatal to the thesis, but it changes the discount math, and markets price timing as much as they price outcomes.

The counter-case, articulated most bluntly from the political side, is that unilateral restraint is not restraint at all — it is a transfer. If American labs slow and others do not, the argument runs, safety is not improved and market share is lost. That framing has proved durable in Washington precisely because it converts a question nobody can settle empirically into one about jobs, investment and strategic position.

The tape is unbothered but not untroubled

Equity markets are treating the exchange as noise rather than news, though the risk tone on the day leans slightly negative. As of the last trade at 18:47 GMT on Monday, Sept. 14, 2026, the S&P 500 proxy SPDR S&P 500 ETF Trust (NYSEARCA: SPY) traded at $762.25, down 0.27% from the prior close of $764.29, inside a session range of $757.93 to $763.52.

The technology-heavy benchmark is doing marginally worse. Invesco QQQ Trust (NASDAQ: QQQ), which tracks the Nasdaq 100, changed hands at $711.93 against a prior close of $714.88, a decline of 0.41%. That is roughly 0.14 percentage points weaker than the broad market, a small gap but the direction you would expect if AI-policy noise were nibbling at sentiment in the most AI-exposed part of the index. The Dow proxy SPDR Dow Jones Industrial Average ETF Trust (NYSEARCA: DIA) was the mildest decliner at $525.04, off 0.14%.

The intraday path matters more than the headline percentage. The Nasdaq 100 proxy printed a session low of $702.74 and had recovered to $711.93 by the time of the last trade — an illustrative bounce of about 1.31% off the low. Whatever unease arrived with the morning had largely been bought back by the afternoon. Investors have seen versions of this argument before and have consistently chosen not to reprice the sector on rhetoric alone.

Anthropic's position is a competitive choice as much as a moral one

Anthropic is privately held, so there is no share price to register the reputational cost of its chief executive's stance. That insulation cuts both ways. It allows the company to advocate for slower, more heavily tested deployment without a quarterly scoreboard punishing it, and it also means the argument's market consequences land on other people's shares — the listed chipmakers, cloud providers and power suppliers whose valuations embed the faster path.

Safety positioning is also a product strategy. Enterprise buyers, regulated industries and government customers are exactly the segments that ask about evaluation, red-teaming and deployment controls before signing. A lab that builds its brand on caution is bidding for that revenue, not withdrawing from the market. Critics reading the call as an attempt to raise rivals' compliance costs are making a commercial accusation, not a scientific one — and it is worth separating the two when weighing the exchange.

The MarketWatch account of the dispute captures how quickly the safety question has been absorbed into the wider political fight over AI, with criticism arriving from the White House and from prominent investors in the same week.

What determines whether this becomes a repricing event

Rhetoric alone will not move AI-linked valuations. Three things would.

• Binding rules. A federal or state requirement for pre-deployment testing, with real enforcement, would put a measurable cost and a measurable delay into every frontier release. That is a modelable change; a speech is not.
• A coordinated industry pause. If more than one leading lab adopted a slower cadence, hyperscaler capital-expenditure guidance would be the first place it showed up, and that guidance is the single most-watched input for the semiconductor and power complex.
• A safety incident that forces the issue. Nothing collapses the distance between the safety camp and the acceleration camp faster than a failure serious enough to make regulation inevitable rather than debatable.

Absent one of those, the fight is likely to stay where it is now: loud, personal and largely uncorrelated with prices. The broader risk for investors is not that development slows. It is that the policy environment becomes volatile enough that the terminal rules are unknowable, which is the condition under which multiples compress even while revenue keeps growing.

For now, the market's answer is visible in the numbers: a modest down day, the AI-heavy index slightly weaker than the rest, and a recovery off the lows well before the close. That is skepticism about the argument's near-term consequences, not endorsement of either side.

Key facts

• Nasdaq 100 proxy (QQQ): $711.93, -0.41%, as of 18:47 GMT Sept. 14, 2026
• S&P 500 proxy (SPY): $762.25, -0.27%, prior close $764.29
• Dow proxy (DIA): $525.04, -0.14% on the day
• Principals in the dispute: Donald Trump and Michael Burry vs. Anthropic's CEO

Frequently asked questions

What did Trump actually say about slowing AI down?

President Donald Trump described calls for a slowdown in artificial-intelligence development as a "sick conspiracy." The remark was directed at the broader push for restraint in frontier model development, which has been championed most prominently by Anthropic's chief executive, who recently argued that AI companies need to adopt better safety measures before deploying new systems.

Why is Michael Burry involved in an AI safety debate?

Burry is among the investors publicly criticizing Anthropic's chief executive over the safety-slowdown argument. His involvement matters because he approaches the AI sector from a valuation-skeptic position rather than a political one, meaning the criticism of the safety camp is now arriving from both accelerationist and bearish directions simultaneously.

Did AI stocks fall on the news?

Not meaningfully. As of the last trade at 18:47 GMT on Sept. 14, 2026, the Nasdaq 100 proxy QQQ was at $711.93, down 0.41%, while the S&P 500 proxy SPY was at $762.25, down 0.27%. QQQ had recovered from a session low of $702.74, suggesting the market treated the dispute as rhetoric rather than a repricing event.

Can investors buy shares in Anthropic?

No. Anthropic is a privately held company and has no listed shares, so there is no public market price that reflects investor reaction to its chief executive's safety stance. Exposure to the debate reaches public markets indirectly, through listed chipmakers, cloud providers, data-center operators and power suppliers whose revenue depends on AI buildout.

What would turn this argument into a real market risk?

Three developments would matter: binding pre-deployment testing rules with enforcement, a coordinated slowdown across more than one leading lab that shows up in hyperscaler capital-expenditure guidance, or a safety failure severe enough to make regulation inevitable. Political rhetoric on its own has repeatedly failed to move AI-linked valuations.

Is a safety-first stance bad for a lab's business?

Not necessarily. Enterprise buyers, regulated industries and government customers routinely ask about evaluation and deployment controls before purchasing. A lab that builds its reputation on caution is competing for that segment rather than withdrawing from the market, which is why critics read the safety argument as partly commercial positioning.

Sources

• Calls for an AI slowdown meet vigorous pushback: Trump calls it a ‘sick conspiracy’ — MarketWatch Top

Photo: David Shankbone · BY-SA 3.0 — source