Closing Keynote : Building Trust in AI Metadata Ecosystems

As AI agents increasingly mediate our interactions with information systems, metadata becomes the critical infrastructure enabling these autonomous systems to understand, navigate, and act upon human knowledge. This keynote examines the evolution from traditional metadata frameworks to agent-ready metadata ecosystems, addressing fundamental challenges in trust, accountability, and human agency. Drawing from recent advances in multi-agent systems and responsible AI frameworks, I will explore how metadata standards must evolve to support not just discovery and interoperability, but also agent decision-making, provenance tracking, and ethical guardrails. The talk will present concrete examples from our work in the areas of agentic AI and responsible AI, demonstrating how metadata can serve as both the foundation for agent intelligence and the mechanism for ensuring these systems remain aligned with human values and societal needs.