Larry’s risky business

Oracle, the decades-old database giant, is making an audacious bet on AI by signing a staggering $300 billion compute deal with OpenAI, a company that currently doesn't turn a profit. To fund this pivot, Oracle has taken $43 billion in debt just in fiscal 2026, moving from its traditional high-margin, low-growth database business to the capital-intensive neocloud sector. The company aims to dominate AI inference—using models to process data—rather than training foundation models, viewing startups' AI offerings as features to integrate into its enterprise suite. Founder Larry Ellison's restless ambition drives this risky strategy, but the company faces the real possibility that OpenAI may default on its payments, putting Oracle's massive infrastructure investment at risk.

Wall Street sees Oracle as a bellwether for the AI boom, since it offers a cleaner AI bet compared to Microsoft's complex business. Oracle's stock price and credit default swaps reflect market sentiment on AI's viability. However, critics warn of a gap between vision and execution, noting Ellison's history of overpromising. If the AI bubble bursts, Oracle's massive debt and reliance on OpenAI's uncertain profitability could lead to severe consequences, making this one of the most high-stakes gambles in tech history.