Key Takeaways
- AI is driving market concentration: Since late 2022, a small number of large AI-focused companies have accounted for a disproportionate share of the S&P 500’s total capitalization, total returns, earnings growth, and capital spending.
- Valuations may be stretched: AI investment has become a key driver of US GDP growth, raising concerns that investor enthusiasm could be outpacing corporate fundamentals.
- Excitement is based on improved labor productivity: While many are concerned that AI may eliminate jobs, the positive argument for AI is that it’s a tool that allows people to do their jobs faster and more effectively, making them significantly more productive.
- Profits remain uncertain: Many companies report limited financial returns from the use of AI so far.
- Diversification remains essential: Staying invested and broadly diversified helps manage risk in today’s concentrated market.
Artificial intelligence (AI) has captured the world’s and the market’s attention. Major stock indexes have climbed to new highs, fueled by a small group of technology giants leading the AI charge. While innovation in AI is real and accelerating, investors are starting to wonder: Is this momentum sustainable, or are we in the early stages of an AI bubble?
A Modern-Day Gold Rush
Economic bubbles often start with a surge of excitement around a powerful new idea, followed by rapid increases in asset prices. When speculation outpaces fundamentals, valuations eventually come back to earth. AI has many hallmarks of this kind of euphoria. There are more than 1,200 AI startups around the world with a valuation over $1 billion, according to CB Insights. Tech giants such as Amazon, Meta, and Microsoft are spending billions to expand their data centers, while headline-grabbing deals from OpenAI, Nvidia, and others seem to arrive almost daily.
According to JPMorgan’s Michael Cembalest, AI-related stocks have accounted for 75 percent of the S&P 500’s returns, 80 percent of earnings growth, and 90 percent of capital spending growth since ChatGPT’s debut in late 2022[1]. That’s a remarkable level of concentration in a single theme.
Recent data also show that AI investment contributed roughly 1.1 percentage points of the 1.8 percent US GDP growth in the first half of 2025[2], surpassing consumer spending as the primary driver of economic expansion. In other words, much of the economy’s strength is now linked to AI-related activity, which is an exciting, but potentially fragile, dynamic.
Today, just ten companies in the S&P 500 comprise 42 percent of the overall index capitalization, an all-time high for concentration. By comparison, at the peak of the “Dot Com” era, the S&P 500’s top ten companies comprised just 29 percent of total capitalization. AI chipmaker Nvidia alone represents 8 percent of the entire S&P 500, and Microsoft and Apple follow with over 6 percent each[3], giving three firms control of nearly a quarter of the index. Nine of the top ten companies are technology/AI related, with only the tenth company (Berkshire Hathaway) not being primarily tech oriented.
Lessons From the Past
This pattern of enthusiasm is reminiscent of the late 1990s, when investors poured money into internet-related companies, convinced they were funding a revolution. At the height of the dot-com boom, companies like Intel and Cisco were among the world’s most valuable. When the bubble burst, both lost roughly 80 percent of their market value, and more than 20 years later, they still haven’t fully recovered those losses. While the internet eventually altered the economic landscape, only a few of the first-movers have remained at the top. That history is a reminder that even truly transformative technologies can experience growing pains along the way.
The Promise, and the Reality, of AI
AI’s potential to reshape industries is enormous. It is being used to accelerate drug discovery, improve logistics, and enhance productivity across countless business functions. But its impact on profits remains uncertain. A recent Stanford report found that 78 percent of businesses use AI for at least one business function[4], yet many are not seeing meaningful financial results. Another study showed that while there is broad adoption of AI technology, more than 77 percent of companies reported little to no bottom-line benefit, with most citing cost savings or revenue improvements of less than 5 percent[5].
Several factors contribute to the gap between AI’s potential and its actual performance. Many companies remain in early adoption stages, struggling to scale pilots due to unclear strategies, limited ambition, or cautious leadership, especially in low-margin sectors where they are wary of costly investments. Heavily regulated industries face additional hurdles from outdated systems and slow approval processes. Even high-value functions like sales, marketing, and engineering often underutilize AI amid employee concerns about job security and skepticism from underwhelming early results.
If the enormous spending on AI doesn’t start translating into stronger earnings more broadly, investor sentiment could shift. The market is heavily concentrated in just a few companies that dominate the ecosystem, including OpenAI, Nvidia, CoreWeave, Microsoft, and Google. These firms are highly interconnected through investments, partnerships, and supply chains. In some cases, there are "circular" relationships where a chipmaker like Nvidia invests in an AI firm, which in turn uses those funds to purchase Nvidia's hardware and services. This creates an ecosystem where money flows among a handful of the same players. If one major player falters, the ripple effects could spread quickly, similar to what we saw in the 2008 financial crisis.
Why This Time Could Be Different
Despite these risks, there are good reasons to believe this may not be a market bubble. Today’s leading AI companies are not speculative dot-com startups that rely upon venture capital to fund their growth; they are large, profitable businesses with significant existing cash flow. The AI boom is based on the building of critical infrastructure, like data centers and powerful chips, which represents a significant economic build-out. More importantly, AI is already being applied across industries, creating tangible efficiencies and new capabilities. That is to say that we are already seeing large adoption of the new technology. If AI adoption and innovation is faster than previous new technologies, then present valuations may be justified.
The primary long-term argument for AI is that AI represents a tool to improve labor productivity, allowing people to do their jobs faster and better than in the past. Similar historical productivity enhancing technologies, such as factory automation, the Internet and robotics technology have driven economic growth. Since labor costs represent a significant portion of corporate expenses, AI’s ability to reduce corporate labor costs by increasing labor productivity may justify high AI valuations.
Staying the Course Through Innovation Cycles
For long-term investors, a prudent approach is neither to go all-in on AI investments, nor to bet against AI by trying to time the next market correction. Instead, it’s about participating in the momentum while managing risk through diversification. While much of the recent market growth has been concentrated in US large caps, it’s wise to diversify into other market segments as a hedge against a possible correction within the most highly valued areas:
- International equities can offer exposure to regions less dominated by AI, providing alternative sources of growth such as manufacturing, infrastructure, and renewable energy.
- Fixed income remains a critical stabilizer in portfolios, helping offset volatility when equity markets fluctuate.
- Real assets, like commodities and real estate, can hedge against inflation and tend to move differently than stocks.
- Defensive sectors such as healthcare, utilities, and consumer staples can provide balance during market turbulence. These sectors are known for their stable cash flows and may hold up better during a market pullback. And perversely, these sectors with higher relative labor costs may benefit more from the AI revolution over the long-term if AI is successful in dramatically improving labor productivity.
The Bottom Line
AI is undeniably transforming the world, but markets often move faster than underlying fundamentals. Whether today’s surge represents the beginning of an AI bubble or the foundation of a lasting technological revolution, one principle remains unchanged: diversification is the most effective safeguard. Innovation cycles will rise and fade, but a thoughtfully constructed portfolio is designed to weather them all.
[1] Eye On The Market, Michael Cembalest, JPMorgan, 9/24/2025
[2] Is AI Already Driving US Growth?, Stephanie Aliaga, JPMorgan, 9/12/2025
[3] S&P Global
[4] Artificial Intelligence Index Report 2025, Stanford University
[5] McKinsey & Company, January 2025 Report