AI Without False Confidence: Grounding Scenario Tests in Real Data
To register for this event please log in to your BCI website profile. If you do not have a website profile, please register here.
Exercising is where this matters most. AI can generate more scenarios, faster, and that speed is genuinely useful. The trap is mistaking volume for assurance. A program can run more exercises than ever and still walk away more confident than it should be, if those scenarios rest on dependency data that no longer reflects how the organization actually runs. Regulators have moved past asking whether a plan exists. Under the Digital Operational Resilience Act, the question is whether recovery can be demonstrated under stress, and exercises that look thorough on stale assumptions will not meet that bar.
Fusion experts will show how to ground AI-generated scenarios in a current, validated model of the business, so the added speed produces real assurance rather than a longer list of exercises nobody can stand behind.
Key Takeaways
- Understand where AI strengthens scenario testing and where it misleads
- Recognize why exercise volume can mask weak assurance
- Ground AI-generated scenarios in current, validated dependency data
- Connect deeper testing to what regulators now expect under DORA
- Apply a disclosure and review discipline to your own AI-assisted work
We look forward to welcoming you!
Speakers:
-
Davis DeRodes
Head of Data Science Innovation, Fusion Risk Management
-
Mandy Leavell
Sr Product Marketing Manager, Fusion Risk Management
More on
