#349 From AI Governance to AI Enablement with Stijn Christiaens, CEO at Collibra

2026-03-05 · Source: DataFramed · Field: Technology & Digital — Artificial Intelligence & Machine Learning, Cybersecurity & Data Privacy, Corporate Strategy & Leadership · Depth: Advanced, extended

Summary

AI governance is rapidly evolving, unlike the more mature field of data governance, with new challenges arising from large language models (LLMs), agents, and multi-agent "swarms." Stijn, co-founder of Collibra, a data and AI governance company, highlights the immediate questions facing teams, such as leadership ownership (legal, security, IT, data, or a new AI role) and establishing standards to prevent tool proliferation. He emphasizes that while governance is often perceived negatively, it is crucial for enabling systems to run well, citing MIT research suggesting a positive correlation between effective data/AI governance and business performance. The discussion covers AI governance failures like chatbots providing incorrect policies or agents with root access causing system damage, and successes from companies like McDonald's and Siemens that integrate controls early. Key differences from data governance include the dynamic nature of AI technology and the evolving scope of traceability, moving from data lineage to agent-to-agent provenance.

Key takeaway

For CTOs and VPs of Engineering navigating rapid AI adoption, prioritize establishing a pragmatic AI governance framework that emphasizes enablement over strict control. Focus on creating transparency through agent registries and classifying AI use cases by risk to allocate scarce governance resources effectively. Foster cross-functional collaboration by actively understanding and addressing the needs of development teams, transforming governance from a "no-sayer" to a strategic enabler that ensures long-term, safe, and performant AI integration.

Key insights

Effective AI governance balances control with enablement, fostering collaboration to safely accelerate AI adoption and achieve positive business outcomes.

Principles

• Visibility and transparency are prerequisites for controlling AI systems.
• AI governance must evolve with underlying technology and use cases.
• Consistent measurement against a chosen maturity framework is key.

Method

Establish an agent registry for transparency, classify AI use cases by risk (e.g., EU AI Act tiers), and prioritize high-risk scenarios for control. Foster cross-functional collaboration by understanding and addressing team needs.

In practice

• Implement an agent registry to catalog models, agents, and use cases.
• Categorize AI initiatives by risk level to focus governance efforts.
• Adopt a pragmatic, consistent approach to AI governance, iterating quickly.

Topics

• AI Governance
• Data Governance
• Large Language Models
• AI Agents
• Risk Assessment

Best for: CTO, VP of Engineering/Data, Executive, Director of AI/ML, AI Architect, Legal Professional

Related on AIssential

• Corporate AI Governance Matters Now More Than Ever — Corporate AI governance is essential for responsible innovation, mitigating risks faster than regulation can.
• Scaling AI With Adaptive Governance — Adaptive AI governance requires matching controls to system type and risk, embedding them into workflows, and treating governance as a learning system.
• #338 The New Paradigm for Enterprise AI Governance with Blake Brannon, Chief Innovation Officer at OneTrust — AI governance must fundamentally transform to manage the rapid, non-deterministic scaling of AI agents in the enterprise.
• ​Building trustworthy AI: A practical framework for adaptive governance — Modern AI agent governance requires adaptive, risk-based models enforced by platforms, not outdated manual processes.
• AI Governance Demystified with Dave Trier — Effective AI governance provides a consistent blueprint, accelerating deployment and mitigating risks across diverse enterprise AI initiatives.
• Four Facets of Responsible Enterprise AI Governance — Effective enterprise AI governance requires integrated, collaborative, dynamic, and human-centric approaches beyond mere regulation.

Open in AIssential →

Editorial summary, takeaway, and curation by AIssential. Original article published by DataFramed.