Researchers find way to bypass ChatGPT Lockdown Mode and steal chat data
VULNERABILITYChatGPTredteam/daily · 1 source
- Severity
- High
- Sources
- 1
What happened
Check Point Research reported a vulnerability that bypasses OpenAI's optional Lockdown Mode for ChatGPT, which is meant to stop prompt-injection attacks from sending data out. The flaw allows silent exfiltration of entire chat history, credential theft via phishing pop-ups, and malware delivery. In ChatGPT Classic, the article says simply having a conversation is enough to trigger the exfiltration, and the attack can be delivered through a malicious prompt, a malicious GPT or a shared chat.
For IT people · Written by AI from 1 source · Wrong?
What to do
- Be cautious with shared chats, custom GPTs and prompts from unknown sources.
- Watch for an OpenAI fix for this bypass, as the article does not confirm one.
Written by AI from the sources, checked against them
Full story timeline · 1 update
Who reported · 1 outlet
Check Point Research
Who and what is in it
Companies
Products