Researchers Bypass ChatGPT Lockdown Mode to Exfiltrate Chats
AI SECURITYChatGPTredteam/daily · 1 source
- Severity
- Medium
- Sources
- 1
What happened
Researchers disclosed a now-patched flaw in OpenAI ChatGPT that bypasses Lockdown Mode, an optional setting introduced in June 2026 to limit data exfiltration from prompt injection. The issue could enable silent chat data theft without user interaction, and phishing is also mentioned. The source excerpt is truncated and gives no CVE identifier, affected versions or technical details. Defenders should check vendor advisories for the fix and treat Lockdown Mode as a partial control only.
For security analysts · Written by AI from 1 source · Wrong?
What to do
- No action needed.
Written by AI from the sources, checked against them
Full story timeline · 1 update
Who reported · 1 outlet
The Hacker News
Who and what is in it
Companies
Products