Cyber Risks · The Hacker News

ChatGPhish Vulnerability Turns ChatGPT Web Summaries Into a Phishing Surface

Summary: Permiso Security disclosed a vulnerability it codenamed ChatGPhish, in which the chatgpt.com response renderer trusts Markdown links and image URLs from a third-party page the assistant has just summarized, auto-fetching those images and surfacing the links as live, clickable elements inside the assistant UI. The technique can leak a victim's IP, User-Agent, and Referer details, render malicious links and fake system-style security alerts, and serve a QR code from an attacker's S3 bucket to bypass desktop URL filters.

What you should know: An employee no longer has to open an attachment or click a suspicious message - simply asking ChatGPT to summarize a page during normal browsing can put attacker-controlled links, spoofed alerts, and QR codes inside an interface your users already trust.

Read the original at The Hacker News ↗

Topics
AI Security
Related