Oracle Adopts AI at Scale, but Faster Coding Doesn't Mean Faster Delivery

Over the past two years, US enterprise software company Oracle spent billions of dollars helping other companies run AI, but it was much slower to fully adopt AI itself. Oracle co-CEO Clay Magouyrk admitted at an internal employee meeting that until last year, the company had not figured out how to make generative AI truly useful across its workforce.

The turning point came in April and May this year, when Oracle began deploying ChatGPT Enterprise and Codex. The company set standards, security controls and internal policies at the same time, reaching an 80% usage rate within three months. AI significantly sped up software development, with developers generating code in about a week instead of the two to three quarters it typically took.

But faster development has not yet translated into faster product delivery. Oracle is still working out how to redesign testing, validation, deployment and release management processes. Oracle was also granted early access to Anthropic's Mythos Preview model for scanning code vulnerabilities, finding more issues in two weeks than in an entire prior year, but about 60% to 70% were false positives.