AI Model Risk — Measurement, Reporting and Mitigation

A 4-week live online masterclass focused on how to measure, report, and mitigate AI model risk by combining AI-specific risk techniques with established model risk management practices.

Dates:

Tue 24 March – Tue 14 April 2026 (one lecture per week)

Format:

Live Online • Total: 16 hours • Certificate: yes

What you’ll learn

Quantitative measurement and reporting of AI model risk
Key accuracy and reliability metrics
Mitigation techniques (including behavioural psychology approaches)
Designing reliable AI-based workflows

Schedule

Week 1

(24 Mar): foundations + reporting + workflows & use cases

Week 2

(31 Mar): measurement (aleatoric vs epistemic risk, metrics, statistical testing)

Week 3

(7 Apr): cognitive biases & psychological effects + mitigations

Week 4

(14 Apr): workflow reliability (grounding/RAG, multistep workflows, voting/crowdsourcing)

Session times

NOTE

Week 1 starts one hour earlier for UK/Europe due to DST; Weeks 2–4 follow the standard schedule.

Week 1

1:00–5:30pm UK / 2:00–6:30pm CET / 6:30–11:00pm IST / 9:00am–1:30pm EDT

Week 2-4

2:00–6:30pm UK / 3:00–7:30pm CET / 7:30pm–12:00am IST / 9:00am–1:30pm EDT