How enterprise GenAI can amplify ransomware risk — and how to contain it

This report explains how generative AI amplifies existing ransomware and cybercrime techniques—speeding reconnaissance, credential theft, data exfiltration, and extortion—while distinguishing two threat models: attackers using AI to improve operations and enterprises deploying AI that can be abused if identities or permissions are compromised. It cites real-world examples of AI-assisted operations, highlights risks like prompt injection and excessive delegated permissions, and recommends six controls (inventory, least privilege, traffic/data controls, monitoring/auditing, containment/recovery, and human/policy approvals) to extend existing cyber resilience to enterprise AI.