We're only about a week into 2026, and AI stocks are back on the move. Fueled by bullish comments by Nvidia (NVDA 0.10%) CEO Jensen Huang at the CES Conference, investors see more gains in store for the high-tech sector.
According the The Motley Fool's 2026 AI Investor Outlook Report, investors continue to be bullish on AI stocks in the new year. Nine out of 10 AI investors plan to maintain or increase their AI stock holdings this year, and younger investors have the most confidence in AI stocks.
Building on that research, I'm also bullish on AI stocks this year, and I think there are two key narratives to pay attention to this year. Keep reading to see my predictions for the AI sector in 2026.
Image source: Getty Images.
1. The bifurcation between AI infrastructure and semiconductors will continue
Concerns about an AI bubble have been recurring for investors as AI stocks move higher, but the AI sector isn't a monolith. There are several categories of companies that are using AI in different ways.
One of those, and arguably the riskiest, is the AI infrastructure sector, made up of companies deriving most of their value from building out AI data centers. The most prominent of these are the neocloud stocks like CoreWeave and Nebius, and legacy tech stocks like Oracle have also planted their flag in the infrastructure camp.
Those stocks have all pulled back significantly from previous highs, indicating that any bubble in those stocks have already been significantly deflated.
The infrastructure sector is higher risk than semiconductors because these companies have to spend billions building out data centers before they can monetize them. CoreWeave and Nebius are losing a lot of money, and Oracle is now free-cash-flow negative after accelerating its infrastructure buildout to meet demand.
Those companies, not semiconductors, also absorb the risk of depreciation in the GPUs they're buying, and the business model, this early in the AI boom, still seems unclear.
By contrast, semiconductor stocks have virtually none of those risks right now. Demand for AI chips is well-established and is outstripping supply, and the chip stocks have the upper hand over the data centers due to the depreciation factor, which favors semiconductors. A rapid depreciation cycle due to upgrades mean data centers will have to refresh their GPU inventory faster, leading to more sales for chipmakers.
This relationship will become clearer in 2026, favoring chip stocks over infrastructure stocks.
Image source: Getty Images.
2. Software stocks will start to emerge as AI winners
Thus far, the winners in AI have largely been chipmakers, as well as infrastructure stocks, until the recent pullback.
However, building out all this infrastructure only makes sense if it's going to be used to run software and the massive spend on hardware is a good indicator that even more money will be spent in AI-related software in the long run.
Palantir has been the most obvious AI software winner so far, seeing revenue growth and operating margin improve in nearly every quarter since 2023, thanks to the launch of its Artificial Intelligence Platform (AIP).
OpenAI and Anthropic, the two leading privately held AI model companies, which provide the brains behind AI software, are generating real revenue now. OpenAI said it would top a run rate of more than $20 billion by the end of the year, and Anthropic gave $9 billion as its run rate target, nearly tripling from the year before.
Many of OpenAI's top customers are software companies, including Shopify, Salesforce, and Datadog, and while the cloud software industry looks poised to capitalize on AI, I think there are smaller, overlooked software stocks that have the best chance of surprising in 2026.
These are stocks like Appian (APPN 2.51%), which specializes in workflow automation and is seeing strong traction for its AI platform; Amplitude (AMPL 2.07%), a digital product analytics company that's introduced a wide range of AI agents, and Figma (FIG +0.16%), the user-experience software company that was punished by Wall Street after it said profits would narrow as it invests in new AI products.
For the AI boom to continue, the software industry needs to find success with AI, and those three stocks have a lot of upside potential if they can make that work.
As we move deeper into 2026, look for chip stocks to widen their lead over AI infrastructure stocks to widen and for software stocks to emerge as AI winners.
