Artificial intelligence (AI) may represent the next major wave of computing, but it’s an investment theme that is not relegated to the information technology sector.
Much of the market activity and most of the stock-price moves related to AI in 2023 and the first half of 2024 have been concentrated among a small group of mega-cap stocks. Yet as more use cases for the application of AI emerge, participation in this long-term investing theme appears to be broadening to companies in a range of industries, and across market cap and capital structures.
According to the managers of several Fidelity large-cap, actively managed funds, a small but potentially growing group of companies in a range of sectors either are beginning to use the technology or plan to use it to gain long-term efficiencies.
“I have found more companies that are using AI to gain competitive advantages, and some of them are in industries that don’t immediately spring to mind,” says Sonu Kalra, portfolio manager of Fidelity® Blue Chip Growth Fund (
Non-tech AI infrastructure plays
For example, the utilities sector began to reach the radar of many investors in 2024 as a path to invest in non-tech AI infrastructure, driving gains for certain stocks in the sector. AI requires significant computing power and a lot of energy to return query results. Absent more energy supply, which can take years to build, electricity prices could rise, potentially boosting the profits of utility companies.
The US Federal Energy Regulatory Commission (FERC) noted in 2023 regulatory filings that US electricity demand could increase by 4.7% over the next 5 years, up from a previous estimate of 2.6%. Independently, several electricity providers also said they expected electricity demand to increase, driven partly by data centers and AI computing.
Like companies in other industries, utilities firms may benefit from implementing AI tools to help manage their businesses. Utilities companies often partner with wind, solar, and hydroelectric energy suppliers. Many must manage relationships with solar panels that produce their own electricity and sell a portion of it back to the grid. This requires far more management of energy supply and demand than in the past, as well as the ability to predict each.
To help with these predictions, the utilities industry—typically not an early adopter of technology—is beginning to embrace AI projects that could help boost the efficiency of existing power infrastructure, manage power purchases, and solve energy transmission problems. Some utilities are also exploring how AI can help customers better manage and predict their power usage. Furthermore, several industrial companies are specializing in power management hardware, including power inverters and intelligent power distribution units, and related software that is in strong demand.
As of May 2024, AI tools were used by 4.8% of all US businesses, with 6.9% intending to use them in the coming months. This reflects a technology that is still early in its development and is not yet poised to contribute broadly to corporate profits or overall economic growth.
Fidelity’s Asset Allocation Research Team recently studied the growth rates of past transformational technologies to gauge to what degree AI could contribute to economic growth. The team concluded that AI could contribute to higher productivity over time, although broad contributions to the economy could still be years away.
Even though overall adoption of AI is still low for all industries, the chart below suggests a few economic segments have embraced the use of AI at more than double the overall rate. This includes professional services. As of May 2024, about 12% of firms in professional services reported using AI, compared with under 3% in manufacturing.
“This is one of the reasons I have focused on finding AI-related investments in services sectors, such as financials and health care, where I believe adoption is among the highest, with room to increase,” Kalra says.
AI in financial services
While it’s unlikely to replace human advice, AI is helping several financial companies take away some types of tedious work. For example, AI is being used by some companies to implement strategies to help reduce foreign-exchange risk in certain portfolios.
AI also can help the operations of financial companies by assisting with the analysis of financial crimes and monitoring regulatory and compliance efforts. Nearly every financial company employs people focused on identifying risks related to money laundering, corruption, and bribery. Several services companies are focused on using AI and machine learning to help financial services companies improve their fraud signals. The benefit of AI is that it can be trained to analyze large data sets quickly that would likely take people much more time to evaluate.
Moreover, a few large financial services companies have developed AI-powered systems that help businesses manage and predict cash flow from clients. These types of systems have reduced repetitive tasks and allowed employees to spend more time on bigger-picture initiatives.
Since many of these technologies have yet to be tested over the long term, early adopters in the financial services industry face risks related to AI systems that may make poor decisions. Companies also face communications challenges and need to explain to customers and prospects how AI is being used if they seek to overcome consumers’ unease with the technology. Already, a few large companies have faced lawsuits related to their use of AI, which may partly reflect the public’s wariness about machines helping with financial decisions.
“Even with these challenges, select financial companies are seeing additive returns on their AI investments, which can help provide efficiencies and competitive advantages longer term,” says Daniel Kelley, portfolio manager of Fidelity® Founders Fund (
AI for health care
Over time, AI promises similar efficiencies in the health care industry. Some companies are using AI to reduce the amount of paperwork and other repetitive tasks conducted by doctors and nurses, potentially giving them more time to spend with patients.
AI also is being used for health scanning and screening. It may conduct better analyses of X-rays, CT scans, MRIs, and sonograms than people in some instances. One reason is that AI tools can leverage very large databases to detect certain abnormalities more efficiently and accurately than the human eye. To date, doctors have not relied solely on AI to make diagnoses—and this may always be the case. However, in some cases, AI is helping doctors come to conclusions sooner than they would otherwise, while helping to limit false-positive test results.
The technology also is being used by health insurance companies to determine the types of assistance and care patients may need and to speed up billing and scheduling. It also may help to find suspicious patterns in insurance claims, thus saving the health care system money over time.
While AI in health care shows promise for making the sector more efficient and improving patient outcomes, there are potential drawbacks, including the potential for false claim denials, privacy breaches, and reliability and trust issues that are common in other industries.
“We’ve already seen a few lawsuits in the industry related to the potential drawbacks of AI,” Kelley notes. “This is an investment risk, but not one that’s kept Fidelity from trying to find health care companies that are using AI to find efficiencies ethically and responsibly.”
AI in communication services
Within communication services, some of the most notable AI innovations are happening among the large social media companies. Several are beginning to integrate chatbots that leverage natural language processing. These integrated chatbots serve a free intelligent AI assistant that can write an email, find a recipe, or plan a trip.
Technically speaking, these built-in chatbots can produce text, images, and computer code based on queries, although their functionality is still being refined and the results they produce are not always reliable. So-called AI “hallucinations” and either wrong or misleading answers are still possible. These chatbots likely will need to be refined further before they can be relied upon consistently for work.
That said, some of the AI innovations are helping communication services companies create value for customers. One of the more impressive uses of AI in communication services is how it can be used to improve digital marketing campaigns by delivering targeted and, in some cases, personalized content. Firms in multiple industries are using AI-powered software that can extract insights and analyze trends to find the type of customers that are most likely to buy specific products and services. While some platforms already employ this type of technology on their online properties (Meta has noted that AI recommendations deliver 30% of posts on Facebook and 50% of posts on Instagram), there’s also specialized software that works across ad platforms and device types to make customized recommendations for ad spending.
“AI for digital marketing looks to be a continued area of innovation and competition, as companies seek to improve their ad-targeting capabilities,” according to Kalra.
What about AI for manufacturing?
The industrials sector is helping to power the buildout of data centers to support the wave of AI-related demand. Industrial companies provide electrical equipment for data centers, cooling systems for chips, and create machines that provide additional power generation for these centers.
Adoption of AI technologies in the industrials sector appears to be slower than in services industries. That said, the industrial sector tends to be an extremely competitive market. The pressure to wring out cost efficiencies to get an edge on the competition eventually could drive faster AI adoption.
“AI can help provide management teams with better data analytics, whereas machine learning can help reduce human error and adapt to project changes without major disruptions,” says Kelley.
AI in energy
In energy, companies also are beginning to use AI to find operational efficiencies for new projects. Several are partnering with AI companies to fine-tune production and reduce downtime. Some of these involve using AI to monitor equipment, thus reducing maintenance expenses and improving reliability. A good example is drilling a well, which has historically been a very manual process. It’s increasingly being automated, with a computer directing more of the drilling. In the future, more of this process may be completely automated.
AI also is being used to help reduce carbon emissions by analyzing data from multiple sources, including drone imagery, infrared sensors, and emissions-control equipment.
With AI driving increased power demand growth, there may be increased demand for fossil fuels, particularly natural gas in the US as power plants will need to run harder and longer than previously expected. New gas-fired power plants will need to be built to support the higher electric load growth. Forecasters are still evaluating how much additional natural gas will be needed later in the decade, but it could be quite significant.
Other business models appear likely
Looking ahead, more AI uses and applications seem inevitable, and some of the startup businesses that are focused on these opportunities could become public companies over time. Some of the trends our research team is watching include companies across a range of sectors and industries that are involved with:
- Autonomous code development
- Internal enterprise data searches
- Jobs typically performed by legal assistants
- Creating custom images from text descriptions
- Classroom tutoring, especially in math and science
- Improving human resources and making it more efficient
Portfolio managers at Fidelity continue to invest based on bottom-up analysis of companies, and AI projects and capabilities represent only a part of managers’ fundamental analyses and stock theses. That said, managers are tracking corporate AI-related investments outside of the information technology sector that appear likely to contribute to a strong return on investment, operational efficiency, cost savings, and/or higher profit margins.
“It’s all about reward potential relative to the risks, and I'm finding non-tech AI-related projects that are helping to meaningfully shift the competitive landscape for specific companies,” Kalra says. “The added benefit is that I have been able to purchase the stocks of some of these companies at what I believe are reasonable valuations, setting up the potential for longer-term gains.”