AI Stocks Erase $600B in 2 Days, Then Nasdaq 100 Hits Record

For investors and market strategists, this two-week AI whiplash is a warning about positioning risk in the most crowded trade of the cycle. A single essay from Anthropic's CEO wiped out $600 billion in two sessions, and a product popularity signal from Meta reversed it in one day. That is volatility driven by narrative, not cash-flow revisions.

The AI trade has just produced one of the most extreme two-week sentiment reversals in years, demonstrating how quickly fear and greed can rotate within the same technology theme. On Sep 12, 2026, Anthropic Chief Executive Officer Dario Amodei published an essay calling for a slowdown in the development of cutting-edge artificial intelligence models. The appeal was quickly echoed by OpenAI chief Sam Altman and SpaceX chief Elon Musk. Because these are leaders of the companies building frontier models, the warning carried weight with investors, seeming to confirm recent industry alarms about existential risks. The first trading session after publication saw AI infrastructure equities sell off hard. From Sep 14 through Sep 15, the Nasdaq 100 Index fell 1.5 percent, wiping out more than US$600 billion in market value. CoreWeave and Lam Research Corporation each dropped more than 9 percent as traders concluded that a slower frontier-model race would reduce demand for GPUs, networking gear, and specialized data-center infrastructure.

“

A single essay from Anthropic's CEO wiped out $600 billion in two sessions, and a product popularity signal from Meta reversed it in one day.

That fear leg evaporated almost as quickly as it appeared. Last week, enthusiasm over Meta Platforms' Muse personal assistant swept through markets. Meta shares jumped 11 percent on Monday, Sep 21, putting the parent of Facebook and Instagram on course for its best month in more than a decade after months of concern that heavy AI spending would not pay off. Arm Holdings surged 17 percent the same day, while Intel and Advanced Micro Devices each gained more than 9 percent. The Philadelphia Stock Exchange Semiconductor Index, or SOX, rose 6 percent on Monday and Tuesday, helping push the Nasdaq 100 to its first record since early June. In two weeks, the market had gone from pricing an AI slowdown to pricing an AI consumer-adoption breakthrough.

The core lesson is that AI equity performance is currently driven less by earnings revisions or capital-expenditure forecasts and more by narrative shifts. Nancy Tengler, chief executive officer of Laffer Tengler Investments, captured the dynamic when she said this has been a narrative market and the narrative changes every other week. The initial selloff in AI infrastructure names such as CoreWeave and Lam Research made intuitive sense: if frontier labs pause or slow the drive toward ever-larger models, demand for the compute that trains those models could flatten. The subsequent rally in Meta, Arm, Intel, and AMD reflected the opposite thesis: that consumer-facing AI products such as Muse can turn AI investment into tangible user engagement and monetization, supporting inference demand across chips and platforms.