Risk, Regulation, and Relationships: Visualizing Global AI Governance with Graphs

Session Track: Knowledge Graphs

Session Time: 05:30 am - 06:00 am November 6

Session description

AI risk management is no longer just a policy issue—it’s a developer concern. As AI systems scale, they face increasing scrutiny under diverse regulatory frameworks across the globe. From the prescriptive requirements of the EU AI Act to the voluntary guidance of NIST and IMDA (Singapore), staying compliant has become a complex puzzle for AI teams and governance professionals alike. In this session, the speaker will introduce a graph-based approach to mapping, comparing, and navigating global AI risk frameworks. By modeling the key principles, obligations, and risk categories of major frameworks as a knowledge graph, developers will gain clarity on where these guidelines align, where they diverge, and how they collectively shape responsible AI practices. You will learn how to use graph technology (e.g., Neo4j) to build a multi-framework risk ontology and generate meaningful insights that support risk-aware development, documentation, and audit-readiness. This session will also highlight how such graphs can evolve into powerful tools for AI assurance and cross-border compliance analysis. Attendees will leave with: - A comparative understanding of major AI governance frameworks (EU, NIST, Singapore, ISO, etc.) - A method for structuring regulatory knowledge using graphs - A practical demo of modeling and querying AI risk categories across jurisdictions