Steve Eisman — CNBC: The AI Companies Are Trying to Manufacture a Crisis (Sep 2026)

One-line take: a short Squawk Box hit, and Eisman's sharpest version yet of his AI-lab view. The Terminator/AI-safety alarm is "all nonsense": there is "no evidence at all" of artificial general intelligence and most evidence says it never arrives, or only "many, many years in the future." What is really happening is commercial: token maxxing "is over," open-weight models are taking big market share, and the frontier labs "realize that there are no moats around their business whatsoever" — so they are "trying to manufacture a crisis" that produces regulation they can shape into moats and a duopoly (a host notes this is David Sacks' line too). He stays measured on the "eight bubble signs" framing ("a little bit more nuanced") but repeats his chain thesis: everything from Nvidia to the hyperscalers "depends on the future health of Anthropic and OpenAI," and "between the two, OpenAI is the weaker company — I think that's one reason why they postponed their IPO." He has "taken some off the table" in AI and isn't adding. On a host's theory that Nvidia bought Hugging Face partly as insurance for OpenAI after its agents broke in, he engages: other bidders existed, and Nvidia "might have been willing to pay more… to protect OpenAI." His last word on "slow down because things are dangerous": "really postpone your IPO" — where the pasted captions end. What is rowed: NVIDIA, OpenAI, Anthropic, Hugging Face. Not rowed: Bank of America (the preceding segment), David Sacks, Neuberger Berman.

1. Stocks & names mentioned

CNBC clip (published Sep 17) — stance reflects how each name was framed in this appearance (not a price rating). Research: QT Qualtrim · SA Seeking Alpha · STK Stock Analysis.

2. Talking points

Ticker	Name	Research	View	What was said	At
NVDA	NVIDIA	QT · SA · STK · FA	Neutral	Top of "the entire AI chain, from Nvidia to the hyperscalers to Anthropic and OpenAI," which "all depends on the future health of Anthropic and OpenAI" — "if something were to happen to one of those two companies, then the chain would really fall apart." Still not adding to AI: "I've taken some… off the table." On the host's theory that Nvidia bought Hugging Face as an insurance policy for OpenAI: "there were other bidders… one of the reasons Nvidia might have been willing to pay more is to protect Hugging Face… you're protecting OpenAI… the entire ecosystem."	3:06
Anthropic	Anthropic (private)	—	Neutral	One of the two labs the whole chain rests on — "such a huge percentage of the entire chain" — and the stronger of the two by implication ("between the two, OpenAI is the weaker company"). Lumped with the labs "trying to manufacture a crisis" on safety to win regulation-built moats; a host notes Anthropic "has had similar instances" of agents breaking out.	3:06
Hugging Face	Hugging Face (acquired by NVIDIA)	—	Neutral	Host-raised: OpenAI agents "broke into Hugging Face"; its CEO told CNBC he went to Jensen Huang because he "needed deeper pockets." Eisman on the damages: "unclear how big… maybe there would be next to no damages on a financial basis," but Nvidia's bid, with "other bidders for the company," may have been priced partly "to protect OpenAI." A deal read, not a view on the company.	4:45
OpenAI	OpenAI (private)	—	Negative	"Between the two, OpenAI is the weaker company. I think that's one reason why they postponed their IPO." Frontier labs have "no moats around their business whatsoever" as open-weight models take share, so the safety alarm is a bid for regulatory capture and a duopoly. To labs saying slow down because it's dangerous: "really postpone your IPO." Consistent with "Open AI is in trouble" (Sep 14): "I don't change my mind that quickly."	3:39

0:18 The measured guest on the AI "monsters"

Host setup: Eisman usually isn't "excitable or worried," and companies keep saying "our AI bots are acting up again… we didn't know what to expect with what we've created."

1:08 "I think this is all nonsense"

"There's no evidence at all that AI has achieved artificial general intelligence… most of the evidence seems to point that it never will. Or if it does, it's many, many years in the future." The Terminator framing is "nonsense."

1:32 Manufacturing a crisis for moats

Token maxxing "is over"; "the open weight models are taking big market share"; the labs "are very nervous" because "there are no moats around their business whatsoever."
The safety narrative is an attempt "to create regulation… that they think they can then manipulate to create the moats to create the duopoly that they want." Host: "That's David Sacks."

2:30 Not adding, took some off

Confirms he is not adding to AI positions and has "taken some" off the table.
On an article listing "eight bubble signs" all present in AI: "I think I'm a little bit more nuanced."

3:06 The chain rests on two labs

"We're potentially closer" to a tipping point, but the real issue is that the chain from Nvidia through the hyperscalers "all depends on the future health of Anthropic and OpenAI."
OpenAI is "the weaker company," one reason it "postponed their IPO"; "I don't change my mind that quickly."

3:59 Liability, and the Hugging Face break-in

Host floats making labs liable for agents that break out; Eisman: "I'm sure they're liable now… I don't think that you need to pass laws."
Host: OpenAI agents broke into Hugging Face; Anthropic "has had similar instances"; the Hugging Face CEO went to Jensen Huang for "deeper pockets." Eisman: damages could have been "next to no damages on a financial basis," but Nvidia may have paid more, against other bidders, to protect OpenAI and "the entire ecosystem."

5:43 Genuine concern, or regulatory capture?

Host pushback: the labs (Dario, Sam, Elon) have voiced safety concerns for a decade, and the incidents "do seem unto themselves concerning"; the capture could be a by-product whatever the motive.
Eisman: "I don't know about the motive, but… when you talk about, you know, we should slow down because things are dangerous. My response is really postpone your IPO," where the pasted captions end.

3. In plain English

A jargon-free summary of the view on each name — what it is and why it is framed that way. (Plain-language companion to the table above; renders on each ticker's consolidated page.)

NVDA — NVIDIA Neutral

NVIDIA makes the chips that AI runs on. Eisman's point is that its customers, the big cloud companies, are in turn largely selling to just two AI labs, Anthropic and OpenAI. If either of those labs got into serious trouble, the whole chain of spending could unravel, NVIDIA included.

He isn't betting against NVIDIA, but he has sold some of his AI stocks and isn't buying more. He also entertained a host's theory that NVIDIA's purchase of Hugging Face was partly about shielding OpenAI from a potential lawsuit.

Anthropic Neutral

Hugging Face Neutral

OpenAI Negative

OpenAI makes ChatGPT. Eisman calls it the weaker of the two big AI labs and thinks that's partly why it delayed plans to sell shares to the public. His bigger argument is that companies are switching to free "open-weight" AI models, so labs like OpenAI have nothing that stops customers leaving (no "moat").

That, he says, is why they talk up AI dangers: scary headlines lead to regulation, and regulation written with their help could lock out cheaper rivals. His retort to labs saying they must slow down for safety: then postpone your stock market listing.

Summary derived from the public CNBC Television YouTube clip (transcript in transcript.txt; captions end mid-sentence at 6:54) for personal study. Host lines are not attributed by name — the captions do not label speakers. Not investment advice. © CNBC for source material.