Trust at Scale - Security and Governance for Open Source Models

Explore the critical security and governance challenges facing organizations adopting open source AI models in this 59-minute MLOps podcast episode featuring Hudson Buzby, Solutions Architect at JFrog. Examine how machine learning has traditionally operated outside conventional DevOps practices and security standards, and discover the new security vulnerabilities introduced by open source model catalogs like HuggingFace and Ollama. Learn about enterprise-scale approaches to addressing security, governance, and legal challenges while enabling developers to safely build and deploy ML/AI applications. Understand the evolution from traditional MLOps to modern platform approaches, analyze the differences between ML and generative AI adoption patterns, and investigate strategies for managing AI tool sprawl within organizations. Gain insights into model trust and safety considerations, organizational debt comparisons, and the value of centralized gateways for AI governance. Discover practical solutions for establishing secure development environments that balance innovation with enterprise security requirements, while exploring the current state of AI adoption failure statistics and their implications for organizational strategy.