Stolen enterprise AI access is being resold through illicit gateways called LLM Mixers, Google’s latest threat report found, according to ZDNET.[1] Google’s threat team said the gateways spread buyers’ requests across stolen accounts to maintain access and reduce detection, ZDNET reported.[1] Mandiant research recorded illicit usage exceeding $100,000 in 24 hours and cumulative charges reaching $1 million at some organizations before access was revoked, according to ZDNET.[1] ZDNET recommended usage limits and alerts for sudden spikes, which can limit damage from compromised accounts.[1] For businesses, that makes controlling AI spending part of containing a security breach.[1]