Luca Pradover - AI Privacy, Security and Best Practices

This session provides a practical overview of how to build and govern AI systems safely, covering: AI governance and ISO/IEC 42001 — the basics of managing AI risk systematically
GDPR and the EU AI Act — what US and international technology companies need to know
Privacy by design — data minimization, retention, RAG, logging, and vendor considerations
AI-specific security risks — prompt injection, sensitive data disclosure, and excessive agency
Secure AI architecture — least privilege, deterministic authorization, isolation, validation, and human approval
Practical best practices for deploying LLMs and AI agents safely in production