Do you know who is using AI as a Service in your Organization? by Gearóid Ó Fearghaíl

This Vectra blog post outlines risks from large language models—such as data exfiltration, fabricated outputs (hallucinations), and lack of source attribution—and describes Vectra’s free ChatGPT Usage Dashboard that monitors DNS requests and uses host attribution to detect and track ChatGPT use for compliance and visibility.