Organizations should resist granting AI more authority simply because its output is fast and confident. 

https://www.securitymagazine.com/articles/102644-reality-check-where-ai-can-be-effective-in-cybersecurity-today