Keep an eye on these three AI stocks

There are reasonably priced stocks that could benefit from additional tailwinds, even though the AI bubble is becoming more and more questioned

Despite growing concerns about their valuations, artificial intelligence (AI) stocks are still very popular and appear to be for some time to come.

The market's top stocks are extremely saturated, and there is growing discussion about the AI megacap bubble bursting. As worries about AI valuations have grown, Nvidia's share price dropped more than 7% during the first week of November, neatly capturing the volatility in this market.

The week "ended very much on the back foot, as AI-focused US tech stocks lost close to £1 trillion in value," according to Fidelity International's investment director Tom Stevenson.

Stevenson, however, dismissed the idea that the AI boom was over. "Talk of another bubble still looks wide of the mark," he remarked. The 50 largest stocks in the SandP 500 saw their valuations double from 20 to 40 between 1998 and 2000, while the remaining 450 stocks' valuations stayed below 20. The difference between the two is much smaller this time29 versus 24. The "

Even if AI proves to be the ultimate bubble, investors can still find value in it.

All of these picks for AI stocks to keep an eye on have some connection to the AI theme in the bull case, and if the AI boom continues to pick up speed as its supporters anticipate, their performance may improve dramatically. However, in the event that the AI bubble bursts, they are either trading at valuations that make them less risky than the headline names, or they have other strings to their bow and aren't totally dependent on AI.

Continue reading to learn about three intriguing AI stocks to keep an eye on as 2025 draws to a close.

Qualcoms inference bid.

I think Taiwan Semiconductor is the best semiconductor stock. Although its forward P/E multiple of 25 is still appealing given its near monopoly on supplying AI giants trading far above it, it is no longer as much of a hidden gem as it once was.

Therefore, Qualcomm (NASDAQ:QCOM) is currently one of the more intriguing semiconductor stocks to keep an eye on, even though Taiwan Semi will always deserve special recognition. Up until recently, it was content to ignore AI in favor of focusing on chips for mobile devices and wireless connectivity.

However, Qualcomm declared on October 27 that it was entering the AI inference market. Two chips, the AI200 and AI250, are anticipated to be released in 2026 and 2027, respectively.

Unlike training, which is the phase where model developers like OpenAI successfully teach AI models how to function, inference is the process by which AI models process prompts and produce responses. Because inference requires more data than training, it could be a profitable market for chipmakers.

The news was well received by the market; on the day of the announcement, Qualcomm shares increased by more than 11%, but they have since declined.

It will be difficult for Qualcomm to compete with companies like Nvidia and AMD in inference chips, but with shares trading at only 14 times forward earnings (as opposed to about 30 times for Nvidia and AMD), expectations are much more measured, and its strong position in wireless chips further reduces the downside risk.

Analysts at Bank of America increased their price target from £200 to £215 in response to the company's most recent results, which is 26% higher than the closing price on November 7.

Clouds of alphabets.

In addition to chips, cloud services are an essential part of the AI infrastructure, which is obviously very profitable compared to (for the time being) deploying large language models (LLMs) themselves.

According to Q3 data from Synergy Research, three cloud service providers hold 62% of the market: Amazon (29%), Microsoft (20%), and Alphabets (NASDAQ:GOOGL) Google Cloud (13%).

In the third quarter of 2025, Google Cloud's revenue rose by 34%, while Microsoft's Azure and Amazon Web Services (AWS) saw increases of 33% and 20%, respectively.

According to Morgan Stanley analyst Brian Nowak, Google Cloud could grow by up to 50% in 2026much faster than the current Wall Street analysts' consensus of 31%. If it can sustain that growth rate, it may eventually overtake Microsoft and even Amazon in terms of market share in the years to come.

Additionally, Alphabet is among the most diverse businesses. It has created its own custom chips for AI and has its own LLM (Gemini), but that doesn't even address its primary search business, which managed to avoid being forced to split up in a September antitrust case.

Despite all of this, Alphabet appears to be significantly undervalued by the market in comparison to the other Magnificent Seven stocks; a forward P/E ratio of 23 seems quite reasonable.

NRG Energy is used in the power play.

One of the challenges facing the AI boom is the high energy consumption of data centers and LLMs. The main obstacles to the development of AI models will probably be the cost and accessibility of energy, particularly in the West.

Therefore, businesses that supply that energy have an obvious advantage. This is already helping diversified suppliers like NRG Energy (NYSE:NRG), whose earnings per share (EPS) rose 32% year over year in its most recent quarter.

Although the stock has nearly doubled so far this year, it is still trading at about 20 times forward earnings. According to analysts surveyed by the London Stock Exchange Group, annual EPS is expected to rise by more than 40 percent between 2024 and 2026.

The author owns stock in Alphabet.

Examine more AI.