Big Short’s Steve Eisman Sells Google, Warns on AI Risk
The investor who saw the 2008 housing collapse coming before almost anyone else is getting cautious again — this time about artificial intelligence.
Steve Eisman, the money manager immortalized in Michael Lewis’s The Big Short, told CNBC he has trimmed his exposure to the AI trade and quietly walked away from one of his longest-held positions: Alphabet, the parent company of Google and YouTube. His message to investors was blunt. If the AI boom fails to live up to the enormous expectations now baked into stock prices, the fallout could be severe.
Here’s what he said, why it matters, and what it might mean for your portfolio.
Eisman cashed out of a stock he’d owned “forever”
Speaking Monday on CNBC’s Squawk Box, Eisman revealed he sold his Alphabet shares a couple of months ago. “I’ve lightened up. I sold my Google a couple of months ago. I’ve owned Google. I can’t even tell you how long I’ve owned Google,” he said, adding that his goal was simply to reduce his exposure to AI. CNBC
That last detail is the important one. This wasn’t a bet against Google’s business. Eisman framed the sale as risk reduction, not a thumbs-down on Alphabet itself. In fact, the company’s numbers have been strong. What spooked him was the bigger picture — how much of the entire market now rides on a single story.
His timing is notable. The move surfaced just as the relentless tech rally took a breather. The S&P 500 and the Nasdaq Composite both logged back-to-back weekly losses, falling 0.6% and 2.1% respectively. CNBC
Why Alphabet, and why now?
The trigger for the broader nervousness wasn’t weak earnings — it was spending.
Last week, Alphabet dropped nearly 8% after raising its capital spending forecast, underscoring the massive investment required to build AI infrastructure. In other words, the company beat expectations on profit and revenue, but investors recoiled at the eye-watering bill for keeping up in the AI arms race. CNBC
The scale of that spending is hard to overstate. Alphabet’s second-quarter 2026 report showed earnings of $9.11 per share against a $3.04 consensus, revenue of $119.80 billion up 24.2% year over year, and Google Cloud revenue of $24.77 billion growing 82%. By almost any measure, that’s a blockbuster quarter. Yahoo Finance
But look at the cost side. Capital expenditures hit $44.92 billion for the quarter, free cash flow turned negative at -$5.86 billion, long-term debt climbed from $46.5 billion to $98.2 billion, and the stock buyback program was suspended. Full-year 2026 capex guidance now sits between $175 billion and $185 billion. Yahoo FinanceYahoo Finance
That’s the catch-22 Eisman keeps pointing to. Investors want to see aggressive AI investment — until they see the invoice. Spend too little and you look like you’re falling behind; spend too much and shareholders panic about returns. Companies are caught in the middle, and so are the people who own their stock.
“It’s all one trade” — the diversification warning
The most striking part of Eisman’s argument wasn’t about any single company. It was about the illusion of safety.
He argued that the classic 60/40 portfolio — 60% stocks, 40% bonds — that millions of Americans hold in their retirement accounts is no longer the diversified cushion people think it is. In his view, AI exposure now bleeds into both sides of that split.
On the equity side, more than half of the stock allocation is effectively tied to tech and AI names. On the bond side, he pointed out that a large share of new bond issuance is itself being raised to fund AI infrastructure. Put those together and the supposedly balanced portfolio starts to look like a concentrated wager on one theme.
“It’s all one trade,” he warned. Even investors who feel diversified, he suggested, may be far more exposed to the AI story than they realize. CNBC
It’s a subtle but unsettling point. Diversification is supposed to protect you when one part of the market stumbles. But if nearly everything you own — stocks and bonds — is quietly leaning on the same driver, then a single shock could hit all of it at once.
What happens if AI disappoints?
Asked directly what would happen if AI failed to deliver, Eisman didn’t hedge. He warned that equity markets would see a “big correction” if the technology does not succeed. Yahoo Finance
He also sees little middle ground. The setup, as he describes it, is increasingly binary: either AI justifies the trillions being poured into it, or the market has badly mispriced the risk. There isn’t much room for a lukewarm outcome.
His caution has been building for a while. Back in December 2025 on the same show, Eisman flagged that performance gains in large language models would begin to slow as they scaled — an observation that made him “nervous,” though at the time he wasn’t selling. By mid-2026, that unease had hardened into action. Crypto Briefing
Eisman also pointed to fresh pressures on the industry. He noted that AI has become far more capital intensive, and that the rise of cheaper AI models from China raises the possibility of a price war. Cheaper competition plus soaring costs is not a comfortable combination for companies betting everything on premium AI returns. Yahoo Finance
He’s sitting in cash — not hiding in “safe” stocks
Here’s a twist that says a lot about the current market psychology. Eisman didn’t rotate his Alphabet proceeds into traditional defensive sectors like consumer staples. Instead, he’s holding cash while he decides what to do next.
His reasoning is telling. He argued that investors today are fixated on a single question — buy AI or don’t buy AI — and simply aren’t interested in rotating into safer, boring corners of the market. “People either want to buy AI or they don’t want to buy AI. They don’t want to shift out of it to buy Clorox,” he said. CNBC
That observation captures the mood better than any chart. In a normal market, money nervous about tech would flow into utilities, household goods, or bonds. In this market, Eisman suggests, there’s nowhere obvious for it to go — which is precisely why the concentration risk feels so acute.
What it means for everyday investors
Eisman was careful to say he is not betting against the broader market. This isn’t a repeat of his famous short position from 2008. He’s not predicting a crash and positioning to profit from it. He’s reducing risk and stepping to the sidelines until the picture clears.
For ordinary investors, the takeaway isn’t “sell everything.” It’s a prompt to actually check what you own. If your retirement account is heavily weighted toward the biggest tech names — directly or through index funds — you may have more AI exposure than you assume. And if a meaningful slice of your bond holdings is financing AI buildouts, the “safe” half of your portfolio may be less independent from the risky half than the labels suggest.
None of this guarantees a downturn. AI could well justify the spending and reward patient shareholders handsomely. Alphabet’s underlying business is, by the numbers, thriving. But Eisman’s core point is about fragility, not doom: when a market leans this hard on one narrative, the margin for disappointment shrinks.
The man who profited from spotting a crowd’s blind spot is once again asking an uncomfortable question — this time, whether investors have quietly turned their entire financial lives into a single trade.