What happened
Announced Oct 27 at Oracle AI World, Oracle AI Security Mesh is a managed platform combining OpenAI frontier cyber models with Oracle's security-testing methods. Specialized AI agents map systems and propose a test plan, investigate weaknesses, and a separate agent independently re-tests to confirm exploitability; it then recommends and helps prepare code fixes, and keeps findings in a private per-customer knowledge base so future tests get smarter.
Why it matters
It is a hyperscaler-delivered, turnkey managed AI security-testing service (application + agent vulnerability discovery and verifiable fix) that lets enterprises surface and confirm exploitable access gaps without building or operating their own AI red-team/pentest capability — a significant new 'Security of AI' workload in the cloud.
Applicability
Enterprise/cloud security and AppSec teams, especially on OCI, evaluating managed AI-driven application and agent vulnerability testing; assess it for production exposure discovery rather than building in-house.