Former White House economists warn AI infrastructure bubble is 'still inflating' and links Anthropic/OpenAI capex to systemic overinvestment risk.

Former White House economists Jared Bernstein and Ryan Cummings said the recent rally in AI-related stocks has strengthened their view that the AI bubble is "still inflating," reiterating their warning from last year that soaring AI valuations increasingly resemble an asset-price bubble.

The economists said the broader stock market has gained about 10% since their warning, while AI-related stocks have climbed around 25%. However, the gap between valuations and heavy spending on the one hand and actual profits on the other has continued to widen.

In a Substack post on Wednesday, they noted that several developments over recent months show that the AI bubble is "alive and well," especially as AI companies prepare to go public. Following SpaceX's record-breaking IPO, AI giants Anthropic and OpenAI have also filed to list their shares. Last month, both Anthropic and OpenAI filed paperwork with the Securities and Exchange Commission to make an initial public offering (IPO), though the price and number of shares to be offered are yet to be disclosed.

Bernstein and Cummings said that until now, investors could only gain limited exposure to these firms through private funding rounds or indirectly via companies such as Microsoft (MSFT) and Amazon (AMZN). They said the public listings do not necessarily signal an imminent market crash. Instead, they will give investors greater transparency to better assess whether current valuations are justified.

"In other words, greater public exposure means more information about profits, losses, share prices, volatility, all factors that will feed into investors' bubble assessments," the post read.

Concerns about an AI bubble are driven mainly by publicly traded technology companies that continue to spend heavily on AI despite limited financial returns. According to estimates from Goldman Sachs, hyperscalers are on track to invest more than $750 billion in AI infrastructure this year and nearly $1 trillion next year. The economists noted that technology investment now accounts for almost 5% of U.S. GDP, exceeding levels seen during the dot-com bubble.

They argued that such spending is difficult to sustain unless it begins to generate meaningful profits. Companies such as Amazon, Alphabet, Microsoft, and Meta have largely funded AI expansion using cash generated from their core businesses, but their free cash flow is shrinking as AI investments continue to rise. Some firms are now turning to debt markets to finance their AI spending.

The economists warned that if AI investments fail to generate strong returns over the next five to seven years, there could be "investor fatigue." That could lead to lower valuations as markets adjust to weaker-than-expected profits, potentially causing the current AI bubble to deflate.