Despite years of investment in AI governance frameworks, data quality programs, and master data management, most organizations are deploying AI agents that operate on fundamentally broken identity data. Fragmented customer, entity, and relationship records mean AI systems are making confident decisions about the wrong version of reality — and nobody is catching it.

This session explores why identity fragmentation is the silent failure mode behind today’s most critical AI governance challenges. We’ll examine how traditional approaches — MDM, CDPs, probabilistic matching — have consistently fallen short, and what the rise of agentic AI demands instead. Attendees will leave with a practical framework for thinking about identity intelligence as foundational infrastructure, not a bolt-on feature, and a clear lens for evaluating where their own governance strategy may have a dangerous blind spot.

Drawing on 20+ years of enterprise data experience — including practitioner research in data governance and decision intelligence — this session is deliberately education-first: you will not see a live demo. You will leave with a mental model you can apply the next morning.

Speaker: Gurpinder Dhillon, Head of Data & AI Ecosystem, Senzing

Dr. Gurpinder S. Dhillon brings over 20 years of enterprise data leadership across data governance, AI enablement, and identity resolution. He holds a Doctorate of Business Administration with a focus on data governance, decision intelligence, and entity resolution — and has spent two decades at the intersection of data strategy and business value, including a senior leadership role at Dun & Bradstreet overseeing one of the company's largest data management business units. He is the author of Think Data, Act AI, and publishes a weekly LinkedIn newsletter, From AI Hype to AI Impact, read by data and AI leaders globally. A frequent speaker at Gartner Data & Analytics Summit, Dreamforce, Big Data & AI World London, and Enterprise Data World, he is known for translating complex data concepts into decisions leaders can act on. His core thesis: AI failures in production are identity failures — not model failures.