Collate 2.0 Demo Webinar Series: Session 3 - Governing Data and AI

​Agents are multiplying faster than anyone is tracking them. When an audit committee asks which AI systems touch financial data, who approved each one, and what each can reach, most teams answer from a spreadsheet that stopped being accurate weeks ago. The underlying problem is structural: data is governed in one system and AI in another, so neither view is ever complete.

​In this session, we'll bring both onto the same graph. We'll show how AI applications, LLMs, MCP servers, and agents are registered as governed assets alongside the data and context they use, how data access is granted with real enforcement back in the source system, and how evidence gets produced when an auditor asks for it rather than assembled after the fact.

​What We'll Cover:

• ​Register and trace – inventory AI assets with owners and stated use, and follow AI lineage from an agent back through its context to the governed data underneath

• ​Review and enforce – route new AI assets through risk council approval, define policies, and check against frameworks including the EU AI Act, NIST AI RMF, and ISO/IEC 42001

• ​Grant and audit – scoped, time-bound data access requests enforced in Snowflake or Databricks, plus the AI Audit Log and machine-readable audit pack export

​AI Governance Studio enters Release Preview with Collate 2.0, available to design partners.