Prerequisite: None
This session will include a moderated Q&A featuring questions from the live audience.
As AI implementation shifts from conversational interfaces to autonomous agentic systems, the "blast radius" of automated actions has expanded exponentially. In this new landscape, governance is no longer a static policy layer, but an active engineering requirement.
In this session, TDWI research fellow Kristy Hollingshead will discuss critical 2026 trends in data and AI governance, moving past the era of "set-it-and-forget-it" data models to the new operational reality of governing systems that use enterprise knowledge assets as live, volatile data sources. Maintaining the integrity of these systems requires a relentless cycle of execution: constant tidying, refreshing, and curating to prevent systemic drift, hallucinations, and bad agentic actions.
Key themes will include:
- Risk mitigation: Strategies for defining and containing the "blast radius" of autonomous agents through rigorous traceability and permissioning.
- Escalation frameworks: Research-based models for placing humans in, on, and over the loop, redesigning how and when systems escalate to human intervention.
- The continuous learning mandate: Why successful governance in 2026 requires moving from static compliance to a model of persistent engagement, including the ongoing "curation tax" required to keep AI knowledge sources reliable and performant.