Last week, Ashley Blake, Principal Product Manager, AI Governance at Collibra and I hosted a webinar on Agentic AI Governance.
The webinar was very well attended and we covered the following topics:
- Collibra AI Governance Module
Ashley presented Collibra AI Governance that integrates seamlessly with the rest of the Collibra platform. - Regulatory Imperatives
AI regulations are multiplying across jurisdictions. Article 5 of the EU AI Act on Prohibited AI Practices goes into effect in February 2025. The U.S. is adopting a sectoral approach to AI regulations including longstanding legislation like the Civil Right Act in addition to state regulations that have already gone into effect. - Human-out-of-the-Loop
Challenges with Agentic AI Governance when agents significantly reduce (or remove) human involvement. - Collibra’s Newly-Released OOTB EU AI Act Assessment
I showcased some recent YDC work on an AI Risk Assessment for LinkedIn Hiring Assistant using the Collibra AI Governance Module. We used the OOTB EU AI Act Assessment from Collibra to assess the use case as High-Risk with associated Deployer obligations. - Stakeholders and Committees
We discussed the use of committees such as the AI Oversight Board and AI Center of Excellence. Ashley highlighted the importance of engaging Procurement and Third-Party Risk Management since the vast majority of AI use cases involve vendor technologies. - Conditional Collibra Workflows Based on AI Score
I presented the use of Collibra workflows where the path is conditioned on the risk rating of the AI use case (e.g., High Risk AI Use Cases are routed to the Operational Risk Management Committee). - Fundamental Rights Impact Assessment
I showcased how the security testing results including prompts and OWASP Top 10 LLM/MITRE ATLAS mappings can be cataloged in the AI use case in Collibra AI Governance.
I will promote the webinar recording when it becomes available. In the meantime, enjoy the holidays and see everyone in the New Year!