The company says it has no evidence the bug was exploited before October’s patch, but researchers say AI agent configuration can still enable prompt-injection style abuse. The post ServiceNow patches critical AI platform flaw that could allow user impersonation appeared first on CyberScoop.