90 days of attacks on AI infrastructure

Wiz Threat Research monitored AI and ML service honeypots over 90 days, detecting persistent attacks exploiting specific AI infrastructure vulnerabilities like remote code execution, prompt injection, and AI-native post-exploitation.