Enabling Adversarial Robustness in AI Models through Kubeflow MLOps

2026-05-14

AI Summary

This work addresses the vulnerability of cloud-native AI models to adversarial attacks during inference by proposing a closed-loop security architecture with attack awareness and dynamic response within the Kubeflow MLOps framework. It pioneers the integration of automated adversarial defense into MLOps pipelines, enabling real-time detection of attacks such as FGSM in Kubernetes environments and automatically triggering defense mechanisms through PGD-based adversarial training. Experimental results demonstrate that the proposed approach significantly mitigates accuracy degradation caused by adversarial attacks and effectively enhances model robustness, thereby offering a scalable security solution for cloud-native AI systems.

Abstract

AI models are increasingly deployed in cloud-native environments to support scalable and automated services. However, while platforms such as Kubernetes provide strong infrastructure orchestration, security mechanisms specifically designed to protect deployed AI models remain limited. This paper presents security measures for AI models deployed in Kubernetes clusters. The proposed architecture integrates Kubeflow-based MLOps to automatically detect adversarial attacks during the inference phase and trigger defense mechanisms that preserve the model's accuracy and reliability. Specifically, a Fast Gradient Sign Method (FGSM) attack is applied at inference time, and a Projected Gradient Descent (PGD)-based adversarial training defense is automatically deployed when a degradation in accuracy is detected. The experimental results indicate that the deployed defense robustifies the model, significantly recovering accuracy relative to the degradation caused by the attack.

Problem

• adversarial robustness
• AI security
• cloud-native AI
• model deployment
• adversarial attacks

Innovation

• Adversarial Robustness
• Kubeflow
• MLOps
• Projected Gradient Descent
• Cloud-native AI