As AI continues to be integrated into business-critical processes, ensuring that models are fit for purpose, reliable, and aligned with organisational values has never been more important. But how do we actually validate AI models in practice, especially when many of us aren’t data scientists?
In this third session of our AI Trust webinar series, we’ll break down the concept of AI model validation in plain language. We’ll demystify the technical processes and show how Risk & Compliance, Internal Audit, and Data Governance functions can play a key role in ensuring trustworthy AI systems.
What we'll cover
Why model validation matters?
- The growing use of AI in operational processes
- Where models go wrong: Lessons from real-world failures
A realistic scenario: validating an AI model
- How do you ensure that your CV Screening model is not introducing bias, is selecting candidates fairly, and is explainable to hiring managers and HR teams?
Deconstructing model validation: what you need to know
- Model validation definition
- Practical examples and means to validate model characteristics (Performance, Fairness, Robustness, Explainability)
Your role in the validation process
- What Risk, Audit, and Control teams should review and challenge
- Common red flags in documentation and governance
- Sample validation Checklist and review template
Moving from silos to collaboration
- How to build a shared language between technical and non-technical teams
- Where validation fits in the model lifecycle (build – validate – monitor)
- Embedding validation into governance structures
- A quick note on GenAI model considerations (to be explored in Webinar Series Part 4)
After the webinar you'll understand
- Understand why validation is more than just a technical checkbox. It’s a cornerstone of AI trust.
- Explore a real-world scenario that makes complex validation concepts easy to grasp and apply.
- Learn how to interpret validation efforts and what questions to ask - even without a data science background.
- Gain practical tools and insights to connect your role to effective model risk oversight.
- Discover strategies to strengthen oversight and collaboration across your organisation.
AI trust webinar series – Part 3: can you trust your AI? A practical guide to model validation
Register for the webinar recording
Contact us
About this webinar series
This session is part of our ongoing webinar series designed to provide you with valuable insights and in-depth knowledge on AI. Each webinar is crafted to build upon the concepts discussed in previous sessions, ensuring a cohesive learning experience.
If you haven't had the chance to attend the earlier webinars in this series, we've made recordings available for your convenience. Simply follow the link to access them and catch up at your own pace.
EU AI Act: key insights for market innovators
Acces the previous webinar here
Providing trust in AI through an ISO42001 certification and Third-Party Assurance attestations
Acces the previous webinar here
