AI security

CISA flags actively exploited LiteLLM AI gateway flaw for federal fix

The command-injection bug, added to CISA's exploited-vulnerabilities catalog on June 8, can be chained for unauthenticated remote code execution and…