Aligning Enterprise AI Security With MITRE ATLAS Using Open Source Technologies

Learn to operationalize AI security in enterprise environments by aligning with the MITRE ATLAS framework using open-source technologies in this 34-minute conference talk. Discover practical strategies for securing AI/ML systems in cloud-native environments as presenters demonstrate how to combine tools like Kubeflow, Clair, Falco, StackRox, and Kubescape to detect and mitigate adversarial threats including data poisoning, model extraction, and evasion attacks throughout the complete AI lifecycle from training to inference. Explore the significance of the MITRE ATLAS framework for AI/ML security, understand how to map open-source tools to the ATLAS matrix for creating actionable layered defenses, and master the integration of security controls into MLOps pipelines using Kubernetes-native tooling. Gain insights into real-world use cases that illustrate effective implementation of AI security measures in enterprise applications where AI systems have become integral to operations.