The artificial intelligence boom has its first serious credibility problem, and it arrived this week in the form of a leaked revenue report that sent tech stocks tumbling.
OpenAI, the company that ignited the generative AI frenzy with ChatGPT in late 2022, is reportedly generating roughly $20 billion less in annual revenue than previously projected to investors. The gap between expectation and reality is not a rounding error — it represents a fundamental miscalculation about how quickly enterprises would pay premium prices for AI tools, and how sticky that spending would prove once the novelty wore off.
The valuation problem
OpenAI's last funding round valued the company at $150 billion, a figure predicated on hockey-stick revenue growth that would justify such stratospheric multiples. The leaked projections suggested the company was on pace for revenues that would support that valuation; the actual numbers suggest something closer to a very good software company than a civilization-altering platform.
The market reaction was swift. Tech-heavy indexes dropped as investors recalculated their exposure to AI-adjacent plays — cloud providers, chipmakers, and the constellation of startups whose business plans assume AI adoption curves that may not materialize on schedule. NVIDIA, which has become a proxy for AI enthusiasm, saw particular pressure.
Enterprise adoption hits friction
The deeper story is about enterprise sales cycles. Companies have experimented enthusiastically with generative AI, but converting pilots into production deployments with serious budgets has proven slower than the breathless conference keynotes suggested. Chief information officers are asking harder questions about return on investment, data security, and whether AI tools actually improve productivity or merely create new categories of busywork.
Microsoft, which has bet heavily on its OpenAI partnership, has reported strong Azure growth but has been notably cautious about breaking out AI-specific revenue. The silence speaks volumes.
Our take
This is not the end of AI, but it may be the end of AI's free pass. For three years, investors have valued AI companies on narrative rather than numbers, funding rounds rather than revenue, and potential rather than profit. OpenAI's stumble suggests the market is finally demanding receipts. The technology remains genuinely transformative; the question is whether the companies building it can capture value at the scale their valuations assume. That question just got a lot harder to answer with a straight face.
