SecurityHacker Newsabout 3 hours ago
Researchers Hack ChatGPT via Shared Links
12 min read
In October 2026, researchers bypassed ChatGPT's sandbox to take over user accounts by sending malicious links. OpenAI fixed the issue on October 26 and classified the report as P1. The attack used a graph and HTML Preview to execute arbitrary JavaScript as the victim, allowing access to private conversations and authenticated APIs.
Level
Hype check
Tap to vote and see what everyone thinks.
In this storyOpenAI
