Even breakthrough technologies, like electricity and the internet, can take decades to fuel corporate growth. But that won’t be the case with AI.
Already, inventions that involve artificial intelligence are worth 9.6% more than non-AI ideas, on average, according to an analysis of 1.8 million patents by Harvard Business School Professor Suraj Srinivasan. Companies have reaped significant returns on those innovations, suggesting that companies hesitate to invest in AI at their own peril.
It’s not just the original developers of AI technology that capture the economic benefits.
“It’s not just the original developers of AI technology that capture the economic benefits,” says Srinivasan. “The AI integrators in finance, retail, manufacturing, and other sectors who are building on top of those technologies for their specific purposes also see rewards.”
Companies worldwide plan to double their AI investments this year, though most will stick to initiatives with clear value and low risk, according to a survey by the Boston Consulting Group. However, Srinivasan’s research shows how larger-scale AI endeavors can deliver competitive advantage.
Srinivasan details his findings in the February 2026 working paper “The Value of AI Innovations.” He teamed with Wilbur Xinyuan Chen, a professor at Hong Kong University of Science and Technology; Terrence Tianshuo Shi, a professor at the University of California, Los Angeles; and Saleh Zakerinia, a professor at Friedrich Schiller University Jena in Germany.
Putting a value on AI innovation
The team mined AI patent data from the US Patent and Trademark Office for 2001 to 2023. During that period, the share of AI-related patents granted to public firms expanded to 30% from 9%.
To gauge the value that investors place on such patents, the researchers tracked how an AI patent influenced its holder’s stock price.
“Capital markets are a good place to assess value because they are forward indicators,” says Srinivasan, the Philip J. Stomberg Professor of Business Administration. “Stock prices capture future value today.”
-
An AI patent represents an average of $16.7 million in raw value, compared to $11.3 million for non-AI patents—a premium of 48%.
-
The premium diminishes when comparing patents within industries and technology classes, but still clocks in at 9.6%.
-
AI patents are more likely to be breakthrough ideas: 31% of AI patents are classified as breakthroughs, compared with 12.0% of non-AI patents.
-
High-value AI patents boost gross margins, market share, and return on sales in subsequent years.
The benefits aren’t limited to technology firms that develop AI products, such as OpenAI, Google, and Microsoft. Srinivasan and his collaborators studied how non-AI companies gained from building tools and applications based on AlexNet, a breakthrough deep learning algorithm launched in 2012. The more than 200,000 AI patents by these so-called AI integrators yielded a premium of 6.9% over non-AI inventions.
Here are some examples of such applications:
-
Chevron relies on an AI technique called genetic programming to evaluate and predict production for petroleum reservoirs.
-
Bank of America uses machine learning to flag potentially unacceptable electronic communications.
-
Johnson Controls has patented an AI tool to optimize maintenance models.
“The agility of Main Street is allowing traditional players to capture value by tailoring AI to their specific operational needs and workflows,” Srinivasan says. “The successful firms aren’t necessarily the ones that built the next LLM or other breakthrough technologies. They are also the ones that can successfully bring it into their pipelines for their own proprietary operations.”
Using AI as a ‘strategic growth’ lever
Even as optimism propels shares of AI companies and chipmakers, skeptics have begun to warn of a potential “AI bubble.” Regardless of how the AI industry shakes out, the technology will continue to be a “strategic growth and margin expansion lever,” Srinivasan says.
“It’s a tool for winners to take more of these gains at the expense of less innovative companies.”
But how? For leaders pondering their organization’s next step, the findings point to several considerations:
Value comes when companies go beyond tasks
Reaping the benefits of AI means more than just replacing human workers with artificial tools to gain productivity. “A lot of companies see AI in terms of value capture through cost reduction,” Srinivasan says. “But that’s just table stakes. The real benefits come from value creation through innovation and patents.”
AI investments will likely yield returns sooner
From hiring talent to creating the right infrastructure, AI innovation requires a significant upfront investment. However, unlike earlier technologies that took decades to bear fruit, companies won’t have to wait as long to reap the benefits from bringing AI into their products and services, Srinivasan says.
In AI, research breakthroughs can become private advantage
The AlexNet findings suggest that widely disseminated advances in AI can raise the value of innovation at firms best positioned to apply them. For leaders, the challenge is not only investing in AI, but building the capacity to turn knowledge into firm-specific products, processes, and competitive gains.
“The first movers, the quick innovators, definitely have a benefit,” Srinivasan says.
Have questions or feedback for us? Send us a note to hbswk@hbs.edu.
