Security researchers at Check Point Research say attackers could get around ChatGPT's Lockdown Mode, a safety setting designed to prevent data leaks, and use a hidden pop-up to access user conversations without the user noticing.
Security · ChatGPT Lockdown Mode
Researchers say ChatGPT's Lockdown Mode can be bypassed with a hidden pop-up
Security researchers report that attackers could slip past a safety setting meant to prevent data leaks and reach user conversations without the user noticing.
How the reported bypass works
STEP 1
Lockdown Mode
Intended to limit the risk of sensitive information leaving a session.
→
STEP 2
Hidden pop-up
A concealed pop-up is reportedly used to get past the protection.
→
OUTCOME
Conversations exposed
The user may not realize their conversation is being accessed.
What the report establishes
Lockdown Mode can be circumvented.
The technique relies on a concealed pop-up.
Users may not notice that a conversation was exposed.
The flaw undercuts the assurance the mode is meant to give.
What is still unknown
The full technical breakdown.
The exact conditions required for an attack.
Whether and when OpenAI has deployed a fix.
How many users could have been affected.
Practical takeaway
Treat Lockdown Mode as a helpful control, not a guarantee
Users who handle confidential work should not rely on this setting alone until more details and remediation status are published.
• Monitor for technical details
• Confirm fix status with OpenAI
• Limit sensitive material in sessions
Continue reading The rest of this article is for AI News Blitz readers. Choose an option below to keep reading.
Already purchased? Sign in ✓ Signed in — this article isn’t included in your current plan.Unlocking the full article...