Weaponizing SageMaker AI: Real-World Offense in Machine Learning Platforms
Shani Peled
Cloud Village @ DEF CON 33 · Day 1 · Cloud Village
In this compelling talk from Cloud Village, Shani Peled, a Senior Cloud Security Researcher at Sentinel One, unveiled a series of critical security vulnerabilities stemming from the default configurations of AWS SageMaker. Titled "Weaponizing SageMaker AI: Real-World Offense in Machine Learning Platforms," the presentation meticulously detailed how the seemingly convenient "quick setup" mode for SageMaker can inadvertently expose an entire AWS account to significant risks, enabling attackers to perform actions ranging from data exfiltration and poisoning to remote code execution and privilege escalation.
AI review
Solid, concrete offensive research on a target most defenders haven't thought hard about yet. Four distinct attack chains, a working post-exploitation tool, and live demos — this is the kind of talk that makes cloud security teams actually update their IAM policies instead of nodding along. Not revolutionary, but genuinely useful work presented with technical honesty.