Chainlit AI framework bugs let hackers breach cloud environments

Published on January 21, 2026

Two high-severity vulnerabilities in Chainlit, a popular open-source framework for building conversational AI applications, allow reading any file on the server and leak sensitive information. [...]