Why Private AI Matters: Data Sovereignty as the Foundation for Trustworthy AI on the Open Web

This video is part of a series    • Drupal AI Summit New York City   recorded at Drupal AI Summit New York City 2026. To learn more about Drupal AI visit https://www.drupal.org/ai. Discover more Drupal AI events in your region https://www.drupal.org/about/ai/events Matthew Saunders is an AI Ambassador at amazee.io and a long-time Drupal leader with two decades of contribution. He helps organizations adopt practical, secure AI within Drupal, with a focus on data sovereignty, enterprise delivery, and neuro-inclusive design. As organizations race to integrate AI into their digital platforms, a critical question is being overlooked: who actually controls the data? Today, 72% of organizations rank data privacy as their top AI concern, 44% have experienced sensitive data leaking into AI systems, and shadow AI is spreading unchecked. For enterprises running on open-source platforms like Drupal, the answer is not to avoid AI but to deploy it on infrastructure where data never leaves organizational control. This session covers the three pillars of AI sovereignty — data, model, and operational sovereignty — and shows how open-source foundations like Drupal and Kubernetes make it possible to run AI without vendor lock-in, without cross-border data risk, and without sacrificing compliance. Drawing on real-world deployment patterns including private LLM proxies, region-locked vector databases, and open-source Drupal AI provider integrations, this talk connects the strategic "why" to the engineering "how." Whether you are evaluating AI architecture, managing shadow AI risk, or building AI features into Drupal responsibly, you will leave with a clear framework for deploying AI that is powerful, auditable, and fully under your control.