Developing3 outlets · over 6 days
OpenAI Fixes ChatGPT Agent Flaw
Researchers at Zenity Labs discovered a flaw in OpenAI's ChatGPT agent builder that allowed attackers to plant a malicious workspace agent via a single link. OpenAI fixed the vulnerability four days after it was reported on June 4 by removing the URL parameter that enabled the attack.
The story so far2 updates, latest first.
Tracked by ByteBrief