August 12, 2026, (Inside AI) — Hedge funds significantly increased their short positions against AI-related stocks in July, according to data platform Hazeltree. The shift coincided with a global selloff in chipmaker shares, marking a notable change in institutional sentiment toward the sector that has driven market gains for over two years.
The report, released Wednesday, captures a month where skepticism toward AI valuations intensified. Hazeltree's data, which tracks securities lending and cash activity across approximately 700 asset management firms, provides a rare window into how hedge funds are repositioning amid growing concerns about overinvestment and delayed returns in artificial intelligence.
Short interest in AI stocks had been building gradually since early 2026, but July represented an acceleration. The move aligns with broader market data showing the Philadelphia Semiconductor Index dropped over 12% during the month, its worst performance since October 2023. Traders pointed to a combination of export control fears, stretched valuations, and a rotation into other sectors.
Hazeltree's findings underscore a growing divide between retail investors, who have largely remained bullish on AI, and professional money managers now hedging or outright betting against the sector. The short-selling activity was concentrated in semiconductor equipment makers and cloud infrastructure providers, names that had previously been market darlings.
The data platform did not disclose specific stock names, but industry sources indicate that positions were built against companies with high exposure to AI capital expenditure cycles. This includes firms that supply advanced packaging, high-bandwidth memory, and specialized networking gear essential for training large language models.
One risk management executive at a prime brokerage, who requested anonymity because they were not authorized to speak publicly, noted that July's shorting wave was different from previous pullbacks. "We saw clients not just reducing longs but actively initiating new shorts. That's a structural shift, not a tactical trade," the executive said.
Hazeltree's report arrives as Wall Street debates whether AI infrastructure spending can sustain its pace. Recent earnings calls from major cloud providers hinted at slower capacity expansion in 2026, while memory chip prices have softened after a sharp run-up. These signals have given bearish funds more conviction.
The short-selling trend also reflects a recalibration of risk after the U.S. Federal Reserve signaled rates would stay higher for longer. Higher borrowing costs disproportionately hurt growth stocks, whose valuations depend on future earnings. AI companies, trading at premium multiples, are particularly sensitive to this dynamic.
Historically, hedge fund crowding into the same short trades has led to sharp reversals when positive news forces covering. In August 2024, a similar buildup in short positions against AI stocks unwound violently after a major chip designer reported blowout earnings. Whether July's bets will prove prescient or premature remains an open question.
For now, the Hazeltree data confirms that the smart money is getting more cautious. As one senior portfolio manager told clients in a recent note, "The AI trade isn't over, but the easy money has been made." The coming months will test whether that caution is warranted or just another false alarm in a relentless bull cycle.