The AI stock selloff that hit global markets on Thursday, October 22, 2026, pushed the Nasdaq and S&P 500 lower and landed hardest on the companies building the data-center boom. An AP-based report carried by the Yakima Herald said the S&P 500 fell 1.4% and was heading toward its worst day since July. The Nasdaq Composite tumbled 2.8% in that count, while other reports put the decline between 2.6% and 2.8%.
The day started in Asia. Newsy Today reported that Asian indexes fell first, with South Korea's Kospi down 3.5%.
Closing figures conflict across outlets, so treat every exact percentage here as unconfirmed until official closing data is checked. One count put the S&P 500 decline at 1.2%, and Nvidia's drop ranged from 3.7% to 4.3% depending on the report.
The selloff has two layers. The first is a financing worry. AI-related stocks have swung on fears that the boom could be threatened by how expensive it has become to borrow money for data centers and AI chips. The second is valuation. Concerns persist that share prices climbed too high during a yearslong run, even though companies in the industry keep reporting growth.
Which AI stocks fell hardest: Nvidia, CoreWeave and power names
Nvidia, the largest US stock by market value, was the heaviest weight on the S&P 500 that day, so its move carried outsized weight in the index.
CoreWeave, which borrows money to buy Nvidia chips and rents them to customers, was among the hardest-hit names. Reported declines ranged from 7.8% to 11%, depending on when the quote was taken. CNN reported earlier this month that CoreWeave carries "tens of billions of dollars of debt," which is why lenders' appetite matters so much to the stock.
Energy equipment companies tied to AI build-outs also fell sharply. CNBC reported that GE Vernova and Bloom Energy led the declines, with each falling more than 11%.
| Name | Reported move |
|---|---|
| S&P 500 | Down 1.4% in the AP-based report, down 1.2% in another count |
| Nasdaq Composite | Down 2.8% in the AP-based report, 2.6% to 2.8% in other reports |
| Nvidia | Down 3.7% in one report, 4.3% in others |
| CoreWeave | Down 7.8% to 11%, depending on the quote |
| GE Vernova and Bloom Energy | Each down more than 11% |
| Brent crude | Above $111 a barrel, highest since May |
Those ranges are the reason to hold off on any decision based on a single quote.
What Jensen Huang's capital comments did to the mood
Nvidia CEO Jensen Huang's remarks on CNBC added fuel to the decline. He said demand for neoclouds is "off the charts" but that capital is scarce and impeding growth, according to CNBC's live coverage. Neoclouds are cloud providers built around rented AI chips, and they depend heavily on outside financing.
Wells Fargo traders said in a client note that his commentary was "spooking markets" and reignited credit-market worries, as Stocktwits reported. That matters because scarce capital hits the same firms that borrow to buy chips.
Meta added to the pressure. CNBC reported that the company's latest earnings comments on the large amount of AI funding it needs were cited as a trigger for shares rolling over. USA 24 has covered how Meta lifted its 2026 AI capital spending floor, which helps explain why investors are watching those numbers.
Funding pressure reaches private companies too. USA 24 reported that OpenAI is nearing a raise of about $30 billion at a $1.25 trillion valuation. Its coverage of Anthropic's draft IPO filing showed $16.3 billion in revenue against an $80.7 billion loss, mostly non-cash, a reminder that the sector's spending runs well ahead of its earnings.
Oil, Iran and the bond market
Oil was the other major driver. Brent crude rose sharply to its highest level since May, with uncertainty tied to the war with Iran, according to BNN Bloomberg. Ground News put Brent above $111 a barrel.
President Donald Trump blamed Iran for the oil surge in a Truth Social post, saying Iran was making up "fake hits." That claim is a political assertion, not a verified finding, so the cause of the oil spike remains disputed. Oil and AI financing worries are both in play, and the data does not show which mattered more.
Government bond yields fell as investors sought safer assets, according to the Washington Times. Falling yields during a stock slide typically signal money moving toward safety, which fits this session.
Markets were already fragile. CNN reported on October 15 that a chip-stock slide had lasted seven straight days after a hot US inflation report raised fears that higher rates would make AI spending more expensive. On the same day, the FTC sent civil subpoenas to OpenAI and Anthropic, a separate policy thread that Engadget reported could matter to AI stock sentiment.
Who is exposed and who may benefit
The broadest exposure belongs to ordinary US investors. Index funds, 401(k) plans and AI-heavy ETFs hold Nvidia and other AI names. Because index funds weight holdings by market value, Nvidia's size gives it outsized weight in many of them.
Data-center developers and utilities that rely on cheap credit face higher borrowing and power costs. Those costs could feed into utility bills and push back data-center plans in the US.
Nvidia's news from GTC Berlin on October 21 cut the other way. The company named six partners for substrate and packaging capacity increases, according to AI Weekly's daily digest. The supply expansion shows the chip supply chain is still growing, even as investors question how the buildout will be financed.
Should you sell Nvidia stock right now?
USA 24 does not give personal investment advice, and the reporting does not support a blanket answer. The data does show concentration risk and a short-term move that may not last. These steps can help you decide:
- Add up your direct holdings in single AI names such as Nvidia, CoreWeave, GE Vernova and Bloom Energy.
- Check the Nvidia weight in your index funds, target-date funds and AI-heavy ETFs, since those funds can hold the same bet without obvious signs.
- Review any leveraged semiconductor exposure. Stocktwits reported that the SOXL semiconductor ETF hit a six-month low during the session.
- Avoid decisions based on one session's quotes. Moves varied by time of day, and closing figures are still unconfirmed.
- Set allocation limits before the next selloff, and confirm them with a licensed financial adviser before making large trades.
What comes next for AI stocks
The next leg depends on three variables: the Iran situation and oil prices, the direction of government bond yields, and whether credit markets keep financing data-center debt on workable terms. If Brent stays above $111 and yields keep falling, investors will likely keep reading this session as a risk-off move. If lenders keep backing neoclouds while chip suppliers expand capacity, the episode may look more like a financing scare than a turning point.
