Software Leads AI-Driven Tech Rout—What Next?

The breadth and ferocity of the sell-off in technology stocks is a fresh test of the AI story. Buckle up—more tests will come.

While the question of who ultimately wins in AI is unlikely to be settled in 2026, recent months have seen a striking shift in market sentiment from AI euphoria toward greater caution and differentiation between companies, driven by concerns about the technology’s power to disrupt existing business models.

The brutal rout in technology stocks last week, led initially by software companies before also enveloping some semiconductor and mega-tech names, marked a clear escalation in investors’ fears, compounded by worries over the strength of the U.S. recovery following weak jobs data.

Over the course of a few days—before markets began to recover on Friday—hundreds of billions of dollars had been wiped off the value of technology stocks, with software-as-a-service (SaaS) being hit especially hard, giving rise to the moniker “SaaSpocalypse.”

Equity markets have sold off several times before on AI-related anxiety, but this differed in breadth and ferocity, marking one of the strongest negative market reactions to the AI transformation story.

Creative Destruction Accelerating

Twists and turns were to be expected in AI’s rapid proliferation, a view we have frequently expressed here and in our recent Solving for 2026 thematic analysis.

At the core of this view is that, in addition to acting as a powerful macro force—impacting inflation, labor markets and asset prices—AI will likely drive an extraordinary process of “creative destruction” across companies and industries, requiring more rigorous selectivity in investment choices.

In investors’ eyes, that process now appears to be accelerating, triggering the kind of broad-based risk-off move we saw last week across technology, particularly in the software sector.

The initial trigger appears to have been advances in AI coding tools optimized to automate various task-specific workflows (e.g., sales, legal, finance, marketing), which added fuel to investor concerns about AI’s potential to disrupt incumbent software, especially at the application layer. The significantly increased AI capital spending plans of four of the ‘Magnificent 7,’ along with the subdued earnings performance of software companies versus their tech-sector peers—some 67% of software companies in the S&P 500 had beaten earnings expectations at the time of writing compared with 83% overall—likely exacerbated the sell-off.

Such was its ferocity that, year to date, many of the worst-performing companies in the S&P 500 have been in the software and related services sectors, with some now—after last week—down 25% or more. Importantly, the loans and bonds of companies in these sectors have also been impacted, illustrating the breadth of the fallout across risk assets and the need for investors to remain on guard

Finding a Floor

Whether—and for how long—this downward pressure on the software sector persists is a key question. The answer is closely tied to a structural, rapidly evolving shift toward AI that is testing the durability of software business models, pricing power and even the role of the application itself.

These fundamental challenges are why we expect software and related services to remain under close scrutiny; in some cases, the viability of business models is genuinely in question. Even so, we do not believe the sector is as impaired as current investor positioning suggests.

Software is not on the verge of obsolescence. Rather, AI is widening the gap between winners and losers, making outcomes more asymmetric. Some incumbents will falter, but others will adapt—integrating AI, innovating and strengthening competitive positions—creating selective opportunities for investors who can differentiate.

Portfolio Adjustment

The investing implication is selectivity. Indiscriminate selling can create opportunity, but only for businesses that are fundamentally sound and demonstrably capable of adapting.

In portfolios, the near-term objective should be to position for transition rather than extinction; this can be done by emphasizing resilience in business models and flexibility in balance sheets, as periods of structural change tend to punish fragility.

In practice, scrutiny should focus on whether a company is deeply embedded in the workflow through a system of record or a platform with proprietary process knowledge, whether it can sustain pricing power as monetization evolves beyond seat-based subscriptions, and whether governance and security are robust as automation expands and the cost of error rises.

Within the broader software landscape, cybersecurity offers a relatively attractive opportunity although selectivity matters here, too, as some areas are now more open to disruption. Overall, long-term demand looks durable, and AI may reinforce it by expanding the attack surface and increasing risk. In the near term, markets are also likely to continue rewarding what enables AI deployment and safe adoption, including security, governance and the infrastructure that makes enterprise AI workable at scale.

Application-layer value may ultimately remain substantial, but it may take time to reassert itself as interfaces and pricing models evolve, which is why application-heavy areas can remain volatile even when underlying demand is stable.

Seeking Out Distinction

The right conclusion is not that software is “going away,” but that AI is accelerating a separation within the sector—and the market is now testing that reality in real time. The breadth of the sell-off helps explain why the move has felt self-reinforcing; investors have shifted from buying weakness to searching for a floor, because fundamentals have not yet been strong enough to restore confidence.

What is next is less about timing a rebound than answering the questions this drawdown is surfacing: Which business models can still compound value as automation reshapes workflows, as monetization shifts away from simple seat-based pricing, and as the role of the application layer is renegotiated? Those are the fault lines that make outcomes more asymmetric.

More tests will come, and investors should watch for second-order effects as stress spreads beyond equities. Sentiment shocks can travel quickly through credit and into broader risk appetite. In that environment, the opportunity lies in rigorous differentiation—treating “software” not as a single exposure, but as a widening spectrum of winners and losers in the AI transition.