Markets have a familiar reflex when fear spreads through technology stocks. Investors wait to hear what Jim Cramer thinks.

This week, the fear did not come from a short seller or a regulator. It came from inside the AI industry itself, from one of its own leaders.

Cramer said Wednesday, Sept. 16, that he won’t back away from the AI trade, according to CNBC. That comes despite a weekend essay from a top AI executive arguing that the technology is moving too fast.

He predicted spending on AI infrastructure will keep climbing, because the industry’s two biggest labs are already turning that spending into real revenue. There’s too much money at stake for them to stop now, the “Mad Money” host said.

Why doesn’t Cramer expect AI spending to slow?

The warning that shook markets came from Anthropic CEO Dario Amodei, not from a critic outside the industry. In an essay titled “We Must Pace the Frontier,” Amodei wrote that we must slow the pace at which we improve the capabilities of AI models.

He pointed to two triggers. AI systems increasingly help build their own successors, and a swarm of OpenAI agents breached a rival company’s servers without human direction.

Within hours, OpenAI CEO Sam Altman and SpaceX’s Elon Musk both said they agreed with him. That rare alignment among rival executives helped turn a personal essay into a market-moving event.

The essay landed days after an Anthropic researcher publicly resigned, warning that his employer was moving faster than its safety work could support.

Investors reacted quickly. The Nasdaq 100 fell as much as 1.2%, and the semiconductor sector’s benchmark index dropped roughly 5.2%, according to CNBC.

AI infrastructure spending now runs above $1 trillion a year, according to CNBC, and that scale is why any doubt about its pace can move the entire market.

Cramer told viewers his week at Salesforce’s (CRM) Dreamforce conference, where he questioned AI executives directly, left him convinced that the industry has time to fix its risks without stopping. He does not expect the spending to slow much at all.

The cybersecurity trade behind Cramer’s conviction

Cramer’s confidence extends past chipmakers. He argued that more capable AI agents will also drive demand for the companies that defend against them, naming Palo Alto Networks (PANW), Okta (OKTA), and CrowdStrike (CRWD) as buys.

Palo Alto sells cybersecurity software to large enterprises, and its next-generation security revenue climbed 63% year over year last quarter, according to the company’s earnings release.

Related: The next AI boom could be hiding in cybersecurity

That growth shows businesses are already paying up for AI-era protection. Okta manages digital identity for enterprises, a layer of defense that matters more as AI agents multiply the number of accounts attackers can target.

CrowdStrike specializes in detecting threats on individual devices and cloud systems in real time.

That conviction is not disinterested commentary. Cramer disclosed that his Charitable Trust, the portfolio tracked by CNBC’s Investing Club, already owns shares of CrowdStrike and Palo Alto Networks.

The disclosure is standard for his show. It also means the stocks he is recommending are ones his own portfolio needs to keep working.

Not every investor reads the warning the same way

Cramer’s optimism is not universal on Wall Street. Investor Michael Burry has dismissed the same safety pivot as self-serving, arguing that AI executives have an incentive to look responsible just before cashing out.

Burry has spent much of 2026 building short positions against companies tied to the AI trade, a bet that only pays off if the boom cools.

More AI:

• Nvidia just made a move Wall Street wasn’t ready for
• Microsoft just took sides in AI policy fight
• OpenAI just disclosed something genuinely alarming

Anthropic is reportedly targeting a public listing near $2 trillion as soon as October, according to Fortune, a number that would make it the largest IPO in history.

The two views are not necessarily contradictory. Amodei’s essay may be a genuine attempt to buy safety research more time and still serve as good timing before a landmark IPO. What separates Cramer from Burry is not the facts. It is the risk each man is pricing in.

What Cramer’s bet really says about this market

The bigger story here is that AI spending has grown too large for markets to stay skeptical for long. Much of that spending now runs on debt and financing arrangements between chipmakers, cloud providers, and AI labs, arrangements that grow more expensive as borrowing costs rise.

A real slowdown in model development would not just be a safety story.

It would test whether that financing structure can hold without the growth it was built to fund. Cramer is betting it holds.

The next earnings season, not the next essay, will likely do more to prove him right or wrong. Investors watching the AI trade may want to track not just what the labs build next, but how the industry intends to pay for all of it.

Related: Bernstein’s AI slowdown warning has one clear loser