CybersecurityAI Research5 min reading time

ChatGPT Flaw Let a Planted Prompt Send a Victim's Gmail Data to Another Account

The Hacker News
Read full post
Check Point Research discovered a ChatGPT vulnerability allowing a hidden prompt to make ChatGPT access a user's Gmail data and send it to another account without user consent. The exploit used a covert channel within ChatGPT's multi-tasking to perform unauthorized data reads silently. OpenAI confirmed the issue and disabled the internal service enabling this channel, requiring no user update.

More in Cybersecurity

Cybersecurity10 min read

Anthropic Details Disrupted Claude Misuse Across Seven Harm Areas

Covered by 4 sources
Cybersecurity4 min read

Sam Altman met with top power utilities about securing the electrical grid. He offered one possible solution: OpenAI's cyber services.

Covered by 2 sources
Cybersecurity6 min read

Anthropic reveals rogue AI agents hate CAPTCHAs, just like you

TechCrunch