Taming the Wild West of ML - Practical Model Signing with Sigstore on Kaggle

Learn how to secure machine learning models and protect against supply chain vulnerabilities through practical model signing implementation using Sigstore on Kaggle. Discover the security threats facing the rapidly evolving ML and LLM landscape, including model poisoning, supply chain vulnerabilities, and challenges in verifying model and data provenance. Explore an efficient solution for signing models with Sigstore at scale, based on real-world integration experience with Kaggle, a leading data science and machine learning platform. Understand the benefits of model signing and master best practices for securing ML workflows. Gain actionable insights that can be applied to other model hubs and platforms to adopt similar security solutions. Focus on protecting ML model integrity through widespread adoption strategies that can prevent significant ML supply chain incidents across the industry.