TechTrending across 3 sourcesThe Register1h ago
Researchers used Claude to hack OpenAI employees' ChatGPT accounts
TL;DRResearchers demonstrated they could breach OpenAI employee accounts using Claude.
Why it matters: Shows AI assistants can execute real attacks; raises security concerns for enterprise AI use.
Agentic exploits for the win (again)
Read full articleSource: The Register · Opens in new tab