Oracle's stock has plummeted 25% in the first half of 2026 as investors grow concerned that major AI customers—particularly OpenAI—may be unable to fulfill multi-hundred-billion-dollar data center contracts. The company's annual report warns of non-payment risks and overestimated customer demand, while Oracle carries $122 billion（約20兆円） in long-term debt and plans to raise another $40 billion（約6.4兆円） to build more capacity.
Summaries like this, in your inbox every morning.
Sign up free →What happened
Oracle's stock fell 25% in the first half of 2026 and is down more than 50% from last year's peak, driven by investor concerns that major AI customers may not be able to pay for contracted data center services. The company's June 10 annual report contained unusually thorough warnings about risks to its AI infrastructure business.
Why it matters
Oracle has accumulated $122 billion（約20兆円） in long-term debt and plans to raise another $40 billion（約6.4兆円） to build more data centers, betting on customer demand. However, the company expects to convert only around 12% of its $638 billion（約100兆円） in remaining performance obligations (contracts signed but services not yet delivered) into revenue over the next 12 months, with a further 34% over the following 24 months—meaning less than half will materialize as revenue over three years in the best case.
What to watch
According to The Wall Street Journal, approximately $300 billion（約48兆円） of Oracle's backlog is attributable to OpenAI alone, which has $25 billion（約4兆円） in annualized revenue, is losing money, and recently raised $122 billion（約20兆円） in funding. The company warned it may experience "risks of non-payment and non-performance" and could end up with infrastructure it cannot use or repurpose.
Oracle built a strong position in AI infrastructure by combining software automation (enabling faster deployment than competitors) with proprietary networking technology (delivering faster data movement than traditional networks). This attracted top AI companies including OpenAI, xAI, and Meta. However, the business model carries a critical structural risk: the company must invest enormous capital upfront—it has taken on $122 billion（約20兆円） in long-term debt and plans to raise another $40 billion（約6.4兆円）—based on customer commitments that may not materialize.
The gap between signed contracts and revenue realization is stark. Oracle's $638 billion（約100兆円） in remaining performance obligations—which surged 363% year over year—represents work promised but not yet paid for. The company's own projections show only 46% will convert to actual revenue over the next three years in the best case, leaving significant risk that much of the infrastructure built will sit idle. This risk is concentrated: a single customer, OpenAI, accounts for roughly $300 billion（約48兆円） of the backlog, yet OpenAI's $25 billion（約4兆円） annual revenue and ongoing losses mean it cannot absorb such obligations without continued external funding.
Oracle's June 2026 annual report explicitly flagged these dangers—warning of "risks of non-payment and non-performance" and noting it could be left with infrastructure it cannot repurpose. For a company carrying massive debt specifically incurred to build this capacity, such a scenario would be catastrophic. Investors have priced in this risk, driving a 50% decline from peak valuation.
No discussion yet for this article
Get curated AI news from 200+ sources delivered daily to your inbox. Free to use.
Get Started FreeFree · takes 30 seconds · unsubscribe anytime
1 minute a day. The AI essentials.
200+ sources · Email / LINE / Slack
