AI Assisted Research Reveals Chained Security Flaws Affecting OpenAI Accounts
Security researchers used Claude Opus 5 to demonstrate a security flaw chain affecting OpenAI employee accounts, with issues reported and fixed responsibly.
CVE-2026-32882
