Course Includes:
- Price: FREE
- Enrolled: 1 students
- Language: English
- Certificate: Yes
- Difficulty: Beginner
This course contains the use of artificial intelligence.
Model risk changed in 2026. SR 26-2 replaced SR 11-7 in the United States, and it places generative and agentic AI outside its scope. The EU moved its high-risk dates. The OCC gave its examiners a new framework for generative and agentic systems. Many teams are still working from older slides.
In this course we follow one fictional bank and its four AI systems: a credit model, a vendor fraud model, an LLM assistant and a refund agent. We build the working papers a model risk team needs: a model inventory, tiering, a validation report, a challenger test, LLM evaluation evidence, a monitoring plan, change management and a committee pack.
You get templates for every paper, four optional free labs that run locally, section quizzes, assignments and a practice test.
Most sections build one working paper, filled in for the bank, so you can see what good evidence looks like before you write your own. The labs use synthetic data and recorded outputs, so you need no customer data, no paid account and no cloud model. We end with a capstone that takes one system from inventory to committee decision.
This course is education. It is not legal or regulatory advice, and it does not promise any supervisory outcome. Check the current sources with your own counsel and compliance team. The bank and the people in the case are fictional.