Wall Street’s relationship with artificial intelligence has entered a new phase. After years in which almost any AI-linked exposure was rewarded, investors are now applying sharper scrutiny, demanding proof of durable earnings power rather than future promise alone. This transition is reshaping performance across sectors, creating a clear divergence between perceived winners and emerging losers as the AI investment cycle matures.

Software Faces a Credibility Test

The sharp sell-off in software stocks over recent weeks reflects growing concern that AI-native tools could erode long-standing subscription-based models. The release of advanced workplace plug-ins by Anthropic intensified fears that general-purpose AI chatbots may replace parts of enterprise software stacks. That anxiety is visible in market performance: the iShares Expanded Tech-Software ETF is down roughly 20% this year, sharply underperforming the broader market.

Still, not all investors agree the repricing is justified. Analysts argue that software companies with deep proprietary datasets and strong customer relationships may be better positioned to integrate AI rather than be displaced by it. The challenge is psychological as much as financial: after years of momentum-driven buying, investors are reassessing risk and demanding clearer paths to monetization. As Steve Sosnick has noted, the market is no longer rewarding vague AI exposure but forcing detailed analysis of competitive advantage.

Hardware Remains the Market’s Conviction Trade

In contrast, confidence in AI hardware suppliers remains robust. Semiconductor makers continue to benefit from the massive capital expenditure plans tied to data centers and cloud infrastructure. The VanEck Semiconductor ETF is up about 13% this year, underscoring investor belief that chips are indispensable regardless of which software platforms ultimately win.

Companies such as Nvidia sit at the center of this narrative, supported by sustained demand for AI accelerators. Memory and storage suppliers have also surged, reflecting expectations that AI workloads will drive exponential data needs. Even as skepticism grows around Big Tech’s ability to earn adequate returns on sprawling data-center investments, the suppliers enabling that buildout continue to attract capital.

Big Tech and the Side-Effect Trade

Large platform companies present a more nuanced picture. Shares of Alphabet and Meta are modestly higher this year, while Amazon and Microsoft have lagged, reflecting investor debate over whether AI spending will translate into profits or simply higher costs.

Beyond direct AI plays, investors are also targeting beneficiaries of second-order effects. Caterpillar, for example, has climbed sharply as demand for data-center construction lifts orders, illustrating how AI’s economic footprint extends well beyond technology.

What Comes Next for AI Investing

As the AI cycle evolves, markets are increasingly categorizing companies into three groups: those that enable the technology, those that successfully adopt it, and those that risk being disrupted. This shift suggests volatility will remain elevated, but it also creates opportunities for disciplined investors willing to look beyond headlines. The era of indiscriminate AI optimism is fading, replaced by a more mature phase where execution, balance sheets, and strategic clarity matter most.