Markets sell off — Nvidia and Oracle slide after OpenAI Q1 miss

AI infrastructure stocks got hit because the market suddenly had to ask a basic question — what if the biggest buyer in the boom isn’t growing fast enough to justify the spending? That was the setup on April 28, when a report said OpenAI had recently missed its own internal targets for revenue and user growth. Oracle, Nvidia, Broadcom, AMD, CoreWeave, and even SoftBank all traded lower as investors repriced the demand story around generative AI. ### What actually spooked investors? The trigger was a Wall Street Journal report, echoed by CNBC and Bloomberg, that OpenAI had fallen short of internal projections for both sales and user growth. That matters because OpenAI isn’t just another software company now — it sits at the center of a giant web of commitments built for a level of demand that takes longer to arrive. ### Why did Oracle get hit so hard? Oracle has become one of the cleanest public-market proxies for OpenAI’s infrastructure appetite. CNBC noted that Oracle has a $300 billion, five-year partnership to supply computing power to OpenAI, so when investors got nervous about OpenAI’s growth, Oracle was one of the first. Oracle’s AI story has become tightly bound to leasing out capacity for customers like OpenAI. ### Why did Nvidia fall too? Nvidia’s drop was smaller — more than 1% — but the logic was the same. Nvidia sells the picks and shovels of the AI boom, so any sign that a major buyer may not monetize usage fast enough can pressure the whole chip complex. Broadcom and AMD also fell, which tells you this wasn’t a company-specific issue. It was a demand-duration scare across the stack. ### Is this really about OpenAI alone? Not really. OpenAI was the spark, but the selloff was about the economics of AI infrastructure more broadly. Bloomberg also flagged weakness in debt tied to data-center companies, including Oracle-linked credit and bonds from SoftBank and CoreWeave. That’s a clue that investors see demand evolving more slowly than expected. ### But didn’t OpenAI just raise a ton of money? Yes — and that’s part of why this landed so hard. OpenAI closed a $122 billion funding round at an $852 billion post-money valuation on March 31. It has also been piling on giant infrastructure commitments, including an expanded AWS deal that adds $100 billion over 8 years only to then see a report about missed internal targets and decided the margin for error may be smaller than it looked a month ago. ### Did OpenAI push back? It did. OpenAI told CNBC the report was “ridiculous” and said management was aligned on buying as much compute as possible. Oracle also defended the partnership and said demand for OpenAI’s technology is still growing quickly. So this is not a case where the companies involved confirmed some collapse in demand. It’s a case where investors saw a crack in the story and immediately tested the most exposed names. ### Why does this matter beyond one bad day? Because the AI trade has been built on a sequence: model demand rises, then cloud usage rises, then chip orders rise, then data-center financing looks justified. The catch is that the spending happens up front, while the revenue proof arrives later. If OpenAI — the company that depends on that curve staying steep. ### Bottom line? This selloff was less about one earnings miss and more about timing. Investors are still willing to believe in AI demand. But they’re no longer willing to assume every dollar of AI capacity spending turns into near-term revenue on schedule.