“Opening the shared link was the only required action”: ChatGPT privacy flaw is fixed

0 points · first seen 21h ago · 1 outlet · 0 comments

A malicious ChatGPT shared-conversation link could have let an attacker read private chats from a signed-in user’s browser session, security researchers say. OpenAI fixed the issue on October 26, according to the research team that reported it. The researchers at Parameter showed that specially prepared content in a shared conversation could run code with the ... “Opening the shared link was… — MIXED

save

· share to X · Bluesky · copy link