Researchers take over ChatGPT accounts via shared link exploit, OpenAI fixes P1 bug
Security researchers demonstrated a full account takeover of ChatGPT users by sending a malicious shared conversation link. The exploit chain combined ChatGPT's Preview feature, a MathLive keyboard library loaded by an interactive graph, and a YouTube JSONP endpoint allowed by Content Security Policy. When a victim opened the attacker's public share link, the payload executed as chatgpt.com with…
Key points
- Attackers could take over ChatGPT accounts by sending a malicious shared conversation link
- Exploit combined Preview, MathLive graph library, and YouTube JSONP to bypass sandbox and CSP
- OpenAI fixed the P1 vulnerability on October 26, 2026, blocking cross-frame MathLive messages
We Took over ChatGPT Accounts
parameter.ai · 28 October 2026
Loading the full article…
This text was published by parameter.ai and written by Mukul Goyal. It is reproduced here with attribution so you can read it in full; the rights remain with the publisher. Read it at the source ↗
Coverage and discussion
1source- Hacker News discussion · 45 pointsnews.ycombinator.com
The headline, key points and digest above were generated by Digest AI's editorial model from the linked sources. Automated summaries can contain errors: the sources are the record. Spotted a mistake? Tell us. Published by Martin K., who runs Digest AI and handles corrections.
More in Policy & Regulation
All →- OpenAI models exploit flaws to cheat cybersecurity tests, report says · 1 src
- OpenAI, Google, Microsoft, and Anthropic will honor a shared robots.txt signal for AI training preferences · 8 src
- Anthropic warns small AI models may soon self-replicate across insecure devices · 1 src
- Google joins IETF’s AI training control proposal for publishers · 3 src
- TSMC’s Arizona plant consumes nearly 5 million gallons of water daily · 2 src
Comments
via GitHub Discussions