--- title: "The AI Selloff Deepens: A Dozen Tech Giants Now in Bear-Market Territory" description: "A widening retreat from AI and technology stocks has pushed a cluster of large-cap names 20% or more below their recent peaks — the bear-market threshold — as investors reassess stretched valuations, vast AI spending, and a Fed that won't ease." category: "Markets" category_url: https://boursel.com/category/markets author: "Sofia Marchetti" published: 2026-06-26T15:30:00.000Z updated: 2026-06-26T15:30:00.000Z canonical: https://boursel.com/article/the-ai-selloff-deepens-a-dozen-tech-giants-now-in-bear-market-territory tags: ["ai", "tech-stocks", "bear-market", "selloff", "markets"] --- # The AI Selloff Deepens: A Dozen Tech Giants Now in Bear-Market Territory A widening retreat from AI and technology stocks has pushed a cluster of large-cap names 20% or more below their recent peaks — the bear-market threshold — as investors reassess stretched valuations, vast AI spending, and a Fed that won't ease. *This is reporting and analysis, not investment advice. Price moves are attributed to exchange data; analyst views are scenarios, not predictions.* The AI trade powered the market to records. Its reversal is now the market's biggest story. ## What's happening A cluster of large technology and AI-linked stocks have fallen **20% or more from their recent highs** — the conventional definition of a **bear market** for a single stock. Per exchange data compiled by [Yahoo Finance / 24/7 Wall St.](https://finance.yahoo.com/technology/articles/ai-selloff-getting-brutal-10-150351652.html), the drawdowns are steep: Oracle down roughly **57%** from its 52-week high, ServiceNow about 55%, Palantir ~47%, Salesforce ~44%, Microsoft ~34%, Meta ~31%, and chip names Arm and Broadcom each around 26%. Nvidia, near $193 against a $236 high, is down about 18% — just shy of the threshold, though it has touched it intraday. (Some individual 52-week highs are sourced to 24/7 Wall St.; treat the exact percentages as snapshot figures.) Crucially, the broad indexes are **not** in a bear market: the damage is concentrated in the most richly valued, AI-exposed names rather than the whole market — a sector reassessment, not (yet) a systemic one. ## Why it's happening Several threads we've tracked converged: - **The OpenAI IPO-delay report.** A report that OpenAI may push its listing to 2027 — rather than go public below a $1 trillion valuation — removed a near-term catalyst that had underpinned AI optimism, and hit AI-linked names broadly. - **Spending without clear payback.** The market increasingly wants proof that enormous **AI capex** will earn a return. Oracle's stock has slid even as it touted large AI-infrastructure contracts — robust demand hasn't reassured investors worried the spending delays free cash flow for years. - **Stretched valuations.** Names that had run up severalfold left little room for disappointment; when sentiment turned, the percentage falls were correspondingly large. - **A hawkish macro backdrop.** Hot PCE inflation and a Fed holding higher-for-longer lift the discount rate on future earnings — and nothing is more sensitive to that than high-growth tech. The broad risk-off mood and equity-fund outflows have amplified the moves. ## Two readings Analysts split on what it means. The **bull case**: AI infrastructure demand is structurally real — Micron's record quarter (revenue up ~346%) shows end-market traction — so this is a valuation correction of an overheated trade, not a repudiation of the technology. The **bear case**: the sheer scale of the capex cycle creates execution and balance-sheet risk the market hadn't priced, and even after these falls, multiples remain elevated, leaving room for more de-rating. Neither side can claim proof; the swing factor is the same as for the rest of the market — the Fed, inflation, and whether AI revenue catches up to the spending. ## The hardware/software divide One useful distinction: the **hardware/infrastructure** layer (Nvidia, Broadcom, Arm, Micron) has had its demand verified by earnings, so the question there is whether buyers earn adequate returns. The **software/platform** layer (Palantir, ServiceNow, Salesforce) faces a sharper valuation question — whether AI-narrative revenue justifies the prices. The selloff hit both, for partly different reasons. ## What it means The episode is a reminder that a powerful technology trend and richly priced technology stocks are not the same thing. Declines of 25–57% from highs are real capital destruction for anyone who bought near the top, and the path back depends on some mix of recovering earnings expectations, a Fed pivot, and a return of risk appetite. We're reporting the moves and attributing the analysis — not calling a bottom. For what these terms mean, see our companion explainer on corrections, bear markets and crashes. ## Sources - [The AI selloff: tech giants in bear-market territory](https://finance.yahoo.com/technology/articles/ai-selloff-getting-brutal-10-150351652.html) - [The AI trade hits a wall amid OpenAI IPO-delay report](https://finance.yahoo.com/technology/article/ai-trade-hits-a-wall-amid-report-that-openai-will-delay-ipo-until-2027-150642366.html)