Three LangChain flaws enable data theft across LLM apps, affecting millions of deployments, exposing secrets and files.
A cyber attack hit LiteLLM, an open-source library used in many AI systems, carrying malicious code that stole credentials ...