Microsoft, Palantir Lead A Software Rout Not Seen Since 2008 Lehman Crisis

Software stocks are witnessing their worst month since the Lehman Brothers collapse.

Investors are now reassessing whether traditional business models can withstand the rapid advance of artificial intelligence. The iShares Expanded Tech-Software Sector ETF (NYSE:IGV) is on pace for its steepest monthly drop since October 2008.

The selloff intensified Thursday, when the sector plunged roughly 6% in a single session — its worst day since March 2020.

The move followed a collapse in Microsoft Corp. (NYSE:MSFT), which dropped more than 12% – the stock's worst day since the lockdown shock.

Indeed, Microsoft beat Wall Street expectations in its latest financial disclosure. But investors instead focused on slowing Azure cloud growth and cautious guidance. This is reigniting concerns about the pace and durability of AI monetization.

High-flying software names like Palantir Technologies Inc (NASDAQ:PLTR), Oracle Corp (NASDAQ:ORCL), AppLovin Corp (NASDAQ:APP) are all down about 20% for the month.

The selloff has traders asking whether this move is technical or something more structural.

Chart: Software Stocks Eye Worst Month Since October 2008

Analysts Question Whether Software Businesses Can Survive AI

Wall Street analysts are increasingly leaning toward the idea that the sector selloff reflects changing fundamentals and not short-term disappointment — as AI begins to rewrite how software is built, sold, and consumed.

Thomas Shipp, head of equity research at LPL Financial, posed a blunt question in a Thursday note: “Can software survive AI?”

"The software business has been one of the best businesses historically," Shipp said. "It is inherently scalable, as once software has been developed, there is a very low incremental cost each time it is sold."

Shipp said subscription-based models have traditionally justified premium valuations. He added those premiums could return if software avoids displacement by AI.

"That said, software companies that are not a system of record do have some risk of displacement," Shipp said.

"Most, if not all, software producers, will need to offer their own AI enhancements to maintain their market share going forward."

Shipp also flagged a longer-term risk.

He said generalized AI tools could slow knowledge worker growth. That could eventually cap software user counts.

A More Radical View: Software Demand May Be Suppressed, Not Disrupted

An even more bearish interpretation on the software business came from Jordi Visser, head of AI Macro Nexus Research at 22V Research.

In a note to clients last week titled "Why Buying ‘Cheap' Software Is the New AI Bubble Trade," Visser indicated that markets are approaching a major inflection point — from a cloud-centric software cycle toward a physical-world infrastructure cycle in 2026.

According to Visser, 2026 marks the shift as AI migrates from centralized data centers into enterprises, devices, vehicles and machines — broadening returns away from software.

"When software can be generated, modified and orchestrated by agents at near-zero marginal cost, ‘build' stops being a strategic project and becomes an everyday action," Visser wrote.

"At that point, the entire buy-side market for software isn't disrupted — it's suppressed."

Visser likened the phenomenon to appetite suppression in patients using GLP-1 drugs. He argues that agentic AI could eliminate "software noise" the same way those medications reduce food cravings.

The implications, Visser said, are philosophical as much as financial. For decades, markets assumed infinite appetite — for calories, stimulation and software. AI challenges that assumption by removing friction, revealing what demand looks like once the triggers disappear.

"The Great Deflation," in this framework, is the market realizing that demand for many high-value industries was inflated by the very inefficiencies they claimed to solve.

The unwind, Visser cautioned, will not happen in a few quarters. But the trajectory, he argues, has shifted.

For investors, the dividing line may no longer be between AI winners and losers — but between companies built for an era of unlimited appetite and those positioned for a world where "enough" is finally defined.

Image: Shutterstock