Concerns over a potential artificial intelligence (AI) bubble have unsettled the stock market recently, raising alarms about broader risks to the U. S. economy. JPMorgan Asset Management reported that a surge in AI spending accounted for roughly two-thirds of U. S. GDP growth in the first half of 2025, surpassing contributions from hundreds of millions of consumers. Major companies have heavily invested in chips and data centers essential for AI operations. A key question persists: Will AI generate sufficient profits to justify the trillions of dollars being spent on its development?Supporters argue that a delay between infrastructure buildout and substantial gains is normal, citing past lags in technologies like the internet. The rapid adoption of AI products such as OpenAI’s ChatGPT reveals huge potential markets, with AI companies focusing on product development rather than profits. Conversely, critics contend that the enormous costs impose pressure for extremely high returns, while so far, little evidence shows that businesses or consumers receive enough value to justify the spending. They stress that AI must prove profitable within a few years, as the current investment pace is unsustainable. Venture capitalist Paul Kedrosky noted it is common for emerging markets to initially lack profit, but AI’s trillion-dollar spending is unprecedented. Both proponents and critics express concerns about economic stakes. David Sacks, a venture capitalist and White House AI advisor, warned a reversal could cause recession, while AI skeptic and NYU professor emeritus Gary Marcus predicted bleak outcomes when enthusiasm fades. The term “bubble” refers to asset prices far exceeding actual value, underscoring doubts about AI’s economic worth and productivity gains. Nvidia exemplifies profit success by selling AI semiconductors, making it the world’s most valuable company, but this reflects demand for AI infrastructure rather than end-user applications.

Currently, AI has not achieved gains proportional to its vast costs. Typically, products generate revenue via direct consumer sales or business usage; AI faces challenges in both areas, with about 95% of AI-investing firms failing to monetize the tech according to a July MIT study, which estimated their combined spend at $40 billion and found limited industry transformation. Consumer profits also lag behind adoption. OpenAI’s ChatGPT, with around 800 million weekly active users, is among the fastest-growing apps ever yet generates far less revenue compared to platforms like Meta, which recently made over $50 billion in three months. OpenAI’s CFO projected roughly $13 billion revenue for 2025, while CEO Sam Altman claimed even higher, accelerating growth and betting on AI clouds, consumer devices, and AI-driven scientific automation to create substantial value. Some analysts highlight rapid chatbot adoption as evidence of AI’s utility and a possible revenue path through ads or paid access. University of Pennsylvania professor Ethan Mollick described it as “the fastest adoption of basically any consumer technology, ” suggesting monetization potential. NYU entrepreneurship professor Arun Sundararajan noted that a delayed business payoff is expected for such paradigm-shifting tech, as experimentation precedes significant productivity gains once firms adapt fundamentally. Others express skepticism due to AI’s heavy infrastructure costs. Unlike typical digital products with low incremental costs, AI’s computational and energy expenses rise with each user prompt, limiting scalability. Kedrosky explained that server maintenance and cooling costs grow with user numbers, posing a profitability challenge. University of Minnesota’s Andrew Odlyzko emphasized the financial strain of trillion-dollar data center investments, which require revenue surpassing even Google’s total to be viable—an enormous hurdle. Despite doubts, some experts affirm AI remains in early development stages, leaving outcomes uncertain. NYU data science professor Vasant Dhar, optimistic about AI’s profit potential, called the current phase “the early innings” and noted that the eventual form and success of AI technology are still unfolding.