AI bubble 'absolutely' here, Lead Edge's Green says capex tops $500B

Lead Edge Capital founding partner Mitchell Green said the market is "absolutely" in an AI bubble, questioning whether record infrastructure spending by hyperscalers will deliver adequate returns on equity.

"We're absolutely in an AI bubble right now," Green said on CNBC's Squawk Box on Aug. 12, discussing the state of the AI boom and the sustainability of capital expenditure programs.

Green's warning lands as hyperscalers and AI infrastructure companies commit record sums to data center buildouts. Ed Yardeni, president of Yardeni Research, separately called the $500 billion Nvidia-Wall Street deal "a little bit of hype," while Steve Eisman, the investor known for the "Big Short," said the future of hyperscalers hinges on whether OpenAI and Anthropic succeed commercially.

The bubble debate carries direct implications for the AI trade. Nvidia, Microsoft, and other AI infrastructure names have driven much of the market's gains, and a reassessment of capex sustainability could trigger significant volatility in tech-heavy indices.

The AI capex cycle has become the defining question for equity markets. Technology companies have committed tens of billions to data centers, GPUs, and related infrastructure, betting that AI adoption will eventually justify the spending. Green's comments suggest that bet may be premature.

U.S. equities posted double-digit quarterly gains in Q2 2026, according to the American Century Focused Dynamic Growth Fund's commentary, with growth stocks broadly outperforming value on optimism around AI infrastructure spending and rising earnings expectations. The fund noted that AI-related stocks, particularly in semiconductors and infrastructure, led gains and shaped portfolio positioning, with emphasis on companies demonstrating "durable AI-driven earnings growth and capital discipline."

The gap between infrastructure spending and realized AI revenue remains wide, even as Nvidia continues to report strong demand for its chips. The American Century fund's emphasis on capital discipline suggests even growth-oriented managers are becoming more selective about which AI names they hold.

Eisman's framing adds another layer. The hyperscaler bet on OpenAI and Anthropic — the two leading AI model developers — means the entire infrastructure buildout depends on those companies converting massive compute investments into profitable products. If either stumbles, the ripple effects would extend across the entire AI supply chain.

Yardeni's characterization of the $500 billion Nvidia-Wall Street deal as containing "a little bit of hype" echoes a broader concern: that financial engineering and speculative enthusiasm may be inflating the AI narrative beyond what fundamentals support.

What the bubble call means for investors

For investors, the question is not whether AI is real — it clearly is — but whether current valuations already price in years of flawless execution. Green's "absolutely" is a direct challenge to that assumption.

The stakes are measurable. AI infrastructure spending has become a significant portion of total technology capital expenditure, and any slowdown would hit Nvidia's data center revenue, Microsoft's Azure growth, and the broader semiconductor supply chain. The American Century fund's emphasis on "capital discipline" suggests even growth-oriented managers are becoming more selective.

Notable Capital's Jeff Richards has identified top AI IPO candidates to watch, suggesting the pipeline of AI companies seeking public listings remains active. But if the bubble narrative gains traction, those IPOs could face a more skeptical reception. The combination of record capex commitments, uncertain monetization timelines, and rising scrutiny from prominent investors creates a fragile setup for the AI trade heading into the second half of 2026.

This article is for informational purposes only and does not constitute investment advice.