Oracle CEO: AI speeds coding but not product delivery

Oracle's AI productivity paradox

Oracle has achieved 80% internal adoption of AI coding tools within three months of rollout, but co-CEO Clay Magouyrk told employees the technology has not accelerated product delivery to customers, according to Business Insider.

The admission exposes a critical gap between individual productivity gains and enterprise-wide results. While AI tools like ChatGPT Enterprise and Codex have made writing code faster, Oracle discovered that speed improvements simply shifted bottlenecks to other parts of the development pipeline—testing, validation, deployment, and release management.

"When you make the actual act of writing the code quicker, it doesn't mean that suddenly everything is 1,000 times faster," Magouyrk said in the internal communication.

Why it matters

Oracle's experience reveals a fundamental challenge facing enterprises deploying generative AI: adoption metrics don't equal return on investment. Companies spending billions on AI infrastructure must redesign entire workflows to capture value, not just accelerate isolated tasks. For Oracle employees producing more code, the workload burden shifts to colleagues responsible for quality assurance and shipping—a dynamic playing out as the company executes layoffs despite a $664 billion backlog.

Layoffs continue amid massive infrastructure spending

Oracle initiated another round of job cuts on September 14, with affected U.S. employees receiving four weeks of base pay plus one week per year of service. California filings identified 441 Bay Area eliminations effective November 13. The company's global headcount dropped from approximately 162,000 to 141,000 during fiscal 2026, a net reduction of 21,000.

The cuts coincide with extraordinary capital expenditure. Oracle spent $55.7 billion in fiscal 2026 and projects $90 billion to $95 billion for fiscal 2027, leaving free cash flow negative by $23.7 billion, as reported by CNBC. The company carries roughly $129 billion in borrowings while racing to build AI infrastructure capacity.

CIO Jae Evans confirmed the rapid internal AI adoption rate, but Magouyrk's assessment suggests Oracle is still determining how to translate tool usage into measurable business outcomes. "When I think about where we were a year ago, I don't think we figured out how we make AI really that useful for ourselves," he told staff.

Stock pressure reflects execution risk

Oracle shares fell nearly 23% through September 11 and closed at $147.61 after another 1.8% decline the following week. The stock trades at 18.13 times forward adjusted earnings, roughly 21% below its sector median, according to Seeking Alpha data.

Analysts expect fiscal 2027 earnings per share of $8.14, up just 6.7%, before a projected 35.1% jump to $11 in fiscal 2028. That acceleration depends on Oracle successfully converting its infrastructure investments and contracted backlog into profitable revenue—an outcome that requires solving the workflow challenges Magouyrk described.

The company's operating margin of 34.27% exceeds its sector average, but capital expenditure equals 105.41% of revenue while gross margin has compressed to 63.96% from a five-year average of 73.03%.

Business Insider first reported details of Magouyrk's internal communication to employees.