Security researchers reportedly breached OpenAI using Anthropic's Claude
TechRadar reports that white-hat hackers breached OpenAI in less than 72 hours. A post citing TechCrunch says they took over employee accounts before accessing an internal code repository.
TLDR
White-hat hackers used Anthropic's Claude to breach OpenAI in less than 72 hours, TechRadar reports. A post citing TechCrunch says researchers used Claude to find and exploit vulnerabilities, took over employee accounts, then accessed an internal code repository. The post says all flaws were disclosed to OpenAI before publication.
Combined views
—
2 Sources, first seen ago
