The Artificial Intelligence Boom Is Not A Bubble—It’s Built To Last

Artificial intelligence (AI) is reshaping markets and driving a surge in capital investment, with its influence reaching far beyond the tech sector. Today, AI-related stocks make up roughly half of the S&P 500, and they also account for about 60 percent of this year’s US market gains. With such rapid growth, it’s natural to ask: are we witnessing another tech bubble forming? 

It’s a fair concern and a risk that we at J.P. Morgan Private Bank are monitoring closely. For now, though, we do not see clear signs of a bubble. AI is not just lifting stock prices—it’s becoming a key engine of the American economy, as well as a foundational driver of long-term productivity. To put it in perspective, real economic growth for the first half of 2025 would be cut in half if all spending on software and information technology (IT) equipment were excluded. Combined with strong earnings and solid forward-looking expectations, the current AI boom appears to be driven by genuine enthusiasm, not unchecked exuberance. 

While the rapid pace of spending may moderate from here, we don’t think the focus on AI is a fleeting trend. For instance, J.P. Morgan Asset Management’s 2026 Long-Term Capital Market Assumptions postulated that technology adoption—especially AI—is expected to provide a near-term boost to profits and a longer-term boost to productivity, helping governments and corporations alike offset demographic headwinds and labor constraints. We anticipate that the benefits will accrue beyond tech and broaden across the economy.  

To that point, the landscape is evolving fast, with companies staying private longer and many leading AI opportunities still emerging. In this environment, we continue to recommend a barbell approach—balancing investments across public and private markets to capture the runway we see ahead. 

So far, we are seeing the evolution of AI unfold in distinct phases. The first phase saw the rise of large-language models powering generative AI. The second phase focused on the hardware and infrastructure required to run these models, spotlighting chipmakers and hyperscale tech companies. We believe we are now entering the third phase: the emergence of “services as a software.”  

In this stage, traditional business services—such as finance, human resources, and customer support—are delivered through intelligent, AI-powered platforms rather than human-led outsourcing or consulting. With this transformation now underway, we are, for instance, seeing AI revolutionizing financial services through automated underwriting and fraud detection, advancing healthcare via diagnostics and drug discovery, and optimizing industrial operations through predictive maintenance and intelligent supply chain management. The “services as a software” opportunity is estimated at US$3 trillion to $5 trillion , with private market assets providing critical access for investors. 

Recognizing that much of the growth in “services as a software” is accessed through private markets, we see industry fundamentals strengthening after the post-2022 correction, with venture and growth investing showing renewed momentum. AI is dramatically accelerating time to profitability—median time to $10 million revenue has dropped from 10 years to just 12 months—while sector-specific dealmaking, especially in AI, is surging, with private market AI deals exceeding $140 billion in 2024 . As exit options broaden and capital scarcity persists, the stage looks set for measured yet meaningful capital deployment into next-generation companies with clearer paths to scale and liquidity. 

As for what we are monitoring as part of our ongoing “AI bubble watch,” we’re tracking several key metrics: 

AI adoption There’s still plenty of room for growth. U.S. adoption is up 60% over the past year, but fewer than 10 percent of U.S. companies are actively using AI in production. The AI adoption cycle is just getting underway, and we anticipate the greatest productivity gains will materialize as AI moves beyond early adopters and becomes a catalyst for change across every industry. 

Power and performance The computing power needed to train cutting-edge AI models is rising exponentially, with performance gains accelerating on both sides of the Pacific. If future breakthroughs demand less power, it could disrupt market leaders who have invested heavily in infrastructure. Still, we believe that broader participation and improved efficiency will fuel wider adoption and drive long-term productivity. 

Capital runway It’s reassuring that major hyperscalers have more profits than debt, and most hold surplus cash. Capex also remains low compared to past overbuilding cycles (such as telecom before the dotcom crash or energy during the shale boom), and grid constraints act as a natural check on spending—while also creating opportunities for utilities, grid operators, and power suppliers. 

Valuations and sentiment Valuations for top AI companies have actually declined relative to the market and past bubbles, again thanks to robust earnings. Recent initial public offering (IPO) returns—often a sign of peak enthusiasm—also haven’t shown a meaningful spike. 

To us, the momentum behind AI looks both strong and well-supported. While history shows markets can sometimes run ahead of technological progress—and this cycle could eventually be no exception—we don’t see bubble trouble for now. Provided that profitability and efficiency remain intact, the AI boom is shaping up to be a lasting engine of growth rather than a transient bubble. As AI platforms and applications evolve, we see compelling opportunities emerging across the AI value chain, across sectors, and in both private and public markets. 

About The Author 

Madison Faller is a Global Investment Strategist at J.P. Morgan Private Bank.