Verizon Report Finds Source Code Leading Data Sent to External GenAI Models as Shadow AI Rises

Verizon's latest breach research has turned shadow AI from a policy concern into a measured insider-risk event stream: source code was the most common data type sent to external GenAI models in its DLP data, while shadow AI rose fourfold to become the third most common non-malicious insider action. The finding puts code, credentials and proprietary technical context at the center of enterprise AI governance.