Between regulatory requirements, risk management, and the pursuit of value, AI governance can no longer remain a purely theoretical exercise. However, many organizations struggle to translate their principles into operational practices. How can we move from statements of intent to governance that is effective, measurable, and tailored to specific use cases? How can we define a framework robust enough to manage risks? What lessons can we learn from early implementations to build sustainable governance? This roundtable will explore the conditions for concrete implementation: trade-offs, organization, processes, responsibilities, performance metrics, and more.