Investment Opportunities in AI: Valuations, Infrastructure, and Growth

Artificial intelligence has become one of the market’s defining investment themes, driving record infrastructure spending, reshaping technology valuations, and prompting investors to rethink where long-term opportunities may emerge. 

But AI is not a single investment story. Opportunities span the entire value chain, from semiconductor manufacturers and data center operators to software providers, consultants, and businesses adopting AI to improve productivity.  

During Morningstar’s latest Investment Opportunities in AI webinar, advisors focused on questions ranging from valuations and infrastructure spending to software disruption and identifying long-term winners. 

Below are the topics that sparked the most interest, along with key insights from Morningstar’s research. 

Perhaps the most common question was whether AI resembles the technology boom of the late 1990s. 

Morningstar Manager Research suggests AI is better understood as an ecosystem than a single investment theme. Companies occupy different positions across the AI value chain and should be evaluated on their individual fundamentals rather than broad market sentiment. 

At the time of the webinar, Morningstar’s Global AI and Big Data Index was trading at roughly an 8% discount to analysts’ aggregate fair value estimates. While that discount had narrowed significantly from earlier in the year, valuations did not indicate a broad market bubble. 

Unlike the dot-com era, many of today’s leading AI companies are generating significant revenue growth and strong margins, supported by growing demand for AI infrastructure. 

Valuation risk is not eliminated, but it suggests the current market is being driven by commercial adoption rather than speculation alone. 

AI infrastructure investment is unprecedented. Approximately $750 billion has been earmarked for AI-related infrastructure spending this year alone, with roughly three-quarters of that investment flowing toward hardware and computing infrastructure, including semiconductors, data centers, networking, power generation, and energy infrastructure. 

One recurring concern was whether companies are overbuilding as major tech firms and model developers form complex capacity-rental deals. However, temporary excess capacity doesn't necessarily signal overspending, as firms must secure infrastructure now or risk falling behind if demand accelerates. 

Looking further ahead, concept proposals like SpaceX’s orbital data centers highlight how aggressively firms are exploring ways to bypass terrestrial power and permitting limits. 

These investments represent a fundamental shift from the asset-light business models investors previously expected. While the transition introduces uncertainty, market performance suggests many investors continue to believe the long-term opportunity justifies the increased spending. 

Morningstar evaluates AI exposure by estimating how much of a company’s future revenue is expected to come directly from AI-related products or services. Manager research complements this analysis by identifying the companies most frequently held across AI-focused investment funds globally. 

Research revealed a surprising trend: every company with a top "core AI" exposure score of 4 was trading at a slight or material discount to fair value—an unexpected finding given market enthusiasm. Conversely, companies with lower exposure scores presented a mixed picture, with over half trading at or above fair value. 

NVIDIA received a score of 3 as an AI supplier, whereas pure-play software providers like Snowflake earned a top score of 4 for creating proprietary AI technology. 

Analysts examined companies that could benefit indirectly from AI adoption. Consulting firms and enterprise service providers could become increasingly important as businesses move from experimentation toward implementing AI at scale. These companies could benefit as organizations seek expertise to integrate AI into existing workflows and operations. 

Looking beyond technology companies to "picks and shovels" businesses supporting AI infrastructure, including industrial and energy companies, may have lost their spark. 

While these businesses remain important beneficiaries of AI investment, many have already attracted significant investor attention and now trade closer to fair value, limiting the valuation advantage they once offered. 

One of AI’s biggest potential disruptors: enterprise software. 

The catalyst for reviewing software moats was the release of Anthropic’s Claude plug-in, which raised concerns that companies might build custom internal applications rather than buy software licenses. 

Morningstar reviewed 132 software companies and downgraded about a third of them. The downgrades reflect heightened uncertainty over how durable competitive advantages will remain over a 10- to 20-year horizon, rather than an immediate loss of switching costs. 

Once products are embedded within an organization’s operations, replacing them requires significant time, investment, and employee retraining. Those competitive advantages remain important, but investors have less visibility into how AI may reshape software business models over the next decade. 

Importantly, there is still little evidence that businesses are abandoning enterprise software altogether. Instead, many organizations appear to be reallocating budgets toward AI infrastructure while evaluating how best to incorporate AI into existing products and workflows.  

Lower-cost AI models, geopolitical competition, regulatory hurdles, and community opposition to new data centers were recurring hurdles mentioned by viewers. 

Our analysts suggest these factors are more likely to influence the pace of AI adoption than fundamentally change its direction. While lower-cost models may improve efficiency, they have not eliminated the need for continued investment in advanced AI infrastructure. 

Likewise, regulatory and permitting challenges may delay some projects, but companies are adapting by investing directly in energy infrastructure and working more closely with local communities.  

As organizations become more disciplined about where AI creates value, investment decisions are increasingly driven by productivity gains and return on investment rather than fear of missing out. 

• AI is better understood as an ecosystem than a single investment theme, with opportunities spanning infrastructure, software, enterprise adoption, and supporting industries.
• Infrastructure spending remains exceptionally high, supported by strong demand for computing power and expectations for continued AI adoption.
• Software companies continue to benefit from durable competitive advantages, but AI has increased uncertainty around their long-term business models.
• Many investors already have meaningful AI exposure through diversified technology holdings, making it important to evaluate how additional AI investments fit within an overall portfolio.
• As AI adoption matures, some of the most compelling opportunities may come from companies that successfully integrate AI into products and services, rather than simply participating in the technology itself.