Book a callBook a callA fixed-scope sprint that inventories every model and AI use, rates the risk and builds your roadmap before OSFI Guideline E-23 takes effect on May 1, 2027. For US institutions, the same sprint maps to the Financial Services AI Risk Management Framework.
Everything an examiner or auditor will ask to see first.
Every model and AI use in one register, including vendor AI, features embedded in the tools you already buy, and the generative AI tools your staff use.
Quantitative and qualitative criteria that set how deeply each model is reviewed, who approves it and which controls apply.
A model risk policy, lifecycle standards and clear roles, from development and approval through monitoring and decommissioning.
A gap assessment against E-23 and B-10, or the FS AI RMF, with a prioritized, owned plan to close each gap before the effective date.
Banks, insurers, and trust and loan companies preparing for E-23, including models supplied by vendors under B-10.
Community and regional banks governing generative and agentic AI that now sits outside formal model risk guidance.
Companies whose credit, fraud or AML models feed a bank partner's inventory and will be asked to prove how they're governed.
Clear phases, named owners and a date on every deliverable.
Weeks 1 to 2: interviews, a review of systems and vendors, and a sweep for unapproved AI tools.
Weeks 3 to 4: every model and AI use recorded, owned and risk-rated.
Weeks 5 to 6: gaps measured against E-23 and B-10, or the FS AI RMF.
Weeks 7 to 8: a prioritized plan to the effective date and a briefing for senior management and the board.
Independent quantitative model validation is delivered with specialist partners and coordinated by OTH Security.