Information security
fromTheregister
2 days agoToo much open-source AI is exposing itself to the web
Exposed, homogenous Ollama open-source AI deployments form a monoculture vulnerable to zero-day exploits, remote compromise, resource hijacking, and unnoticed abuse.