Our US correspondent has another report from a recent wealth sector conference in Chicago, and this article concentrates on the dominant AI theme.

Think Apple is an AI play, as the company hopes investors do?

Think again, said Morningstar principal Kenneth Lamont pictured), speaking at the research firm’s annual Investment Conference in Chicago last week.

In fact, Apple, along with well-known companies like Intel and Qualcomm, are “false friends” of AI investors, according to Lamont, research manager for Europe, Africa and the Middle East based in London. These stocks, along with others like Palo Alto Networks and Datadog, appear frequently in AI-focused portfolios, but aren’t rated at all by Morningstar’s AI equity research team, according to a new Morningstar research report, Opportunities in Artificial Intelligence.

Apple is held in 38 per cent of AI fund portfolios, yet Morningstar’s analysts give it a score of zero. “AI makes up a very low portion of Apple's total sales, given they have a limited offering and it’s just part of iPhone sales,” Lamont said.

AI stocks under the radar
Conversely, the report revealed some “hidden gems” for AI investors, including technology producers Oracle and Snowflake and semiconductor companies Advanced Micro Devices, Broadcom and Taiwan Semiconductor.

Not surprisingly, well-known tech companies like Nvidia and Microsoft were also rated highly by Morningstar analysts, but flying under the radar, with lower exposure but attractive valuations, were software companies HubSpot and ServiceNow along with Cognizant Technology Solutions, an information technology firm and Meta Platforms, an Internet content and information company.

The case for funds
The Morningstar report also makes the case for investing in AI-themed funds, like highest rated First Trust Bloomberg Artificial Intelligence ETF, which features strong access to core AI stocks and selection and weighting driven by direct AI revenue.

Leadership and dominance can change quickly in fast-moving sectors like AI, and diversified thematic funds can increase the chance to participate in the returns of future “shooting stars” without needing to predict them in advance, according to the Morningstar report.

Buying into a bubble?
Investors who are understandably excited about artificial intelligence-themed stocks and funds but worried that they’re buying into an AI bubble shouldn’t be, Lamont said.

Defining a bubble as an asset that has been dislocated in price from its fundamental value, Lamont said AI-themed stocks and funds have been fairly valued and “pretty well priced.” 

Massive spending on AI infrastructure supports future growth, and AI disruption has led to “moat downgrades” at competing companies and sectors that can produce increased revenues and value for AI insurgents, Lamont said.

High AI exposure but undervalued
Indeed, market turbulence this year has left AI-themed investments undervalued, according to the Morningstar research report. Most stocks with meaningful AI exposure are undervalued and stocks with the highest level of AI exposure “are all undervalued,” according to the report.

There are similarities to the early days of the dot com bubble and investing in AI today, Lamont acknowledged. But the “key difference” is that there is “real revenue” and value attached to AI companies, which Morningstar’s equity researchers define as being either a producer or supplier of artificial intelligence, with 50 per cent or more of companywide revenue driven by AI for the next five years and a net profit increase over the same period from exposure to AI.

The Morningstar report outlines bull and bear cases to be made for investing in artificial intelligence.

Bear case for AI
--  AI is energy-intensive and geopolitical shocks such as the Iran war risk structurally higher or volatile power costs, pressuring margins and deployment, according to Morningstar.
--  “AI is a scale game,” Lamont said, and if returns to scale weaken, overcapacity and lower returns on invested capital follow.
--  While AI adoption may be strong, monetization is uneven. Open-source and rising competition may increase usage while compressing pricing power.
--  There is also regulatory risk: current frameworks focus on data/consent, not usage. Tighter rules on training data, liability, or deployment could raise costs and cap the total addressable market.

Bull case
--  AI enables a step-change in productivity, automating cognitive tasks across sectors – not just tech – expanding the addressable market for enablers and platforms.
--  Model performance continues to improve as computational power, data efficiency, and architectures advance, while inference costs fall – supporting sustained investment and broadening adoption.
--  Data centers, semiconductors, networking, and power remain in the early innings of a multiyear capital expenditure cycle, driving durable demand for high-barrier “picks-and-shovels” players.
--  AI is now treated as critical national infrastructure, with coordinated public and private investment, reducing the risk of a cyclical slowdown.