← registry

Researchers used Claude to hack OpenAI

Researchers demonstrated a security vulnerability by using Claude to gain unauthorized access to an OpenAI employee account and sensitive GitHub repositories.

Categorymisuse
Severityhigh
AI systemchatbot
Sectorstechnology
Harm typessecurityprivacy
Lifecycle stageoperation
Actorend_user
Published2026-09-18 13:30:12

Summary is Secursion's own; full text lives at the source. Attribution preserved.