GRA · AI and Financial Model Risk
Spatial financeWhere capital meets hazard, exposure and transition
Resilience and finance readiness
Model risk readiness asks whether a model used in a financial decision, from credit scoring to catastrophe pricing, is inventoried, documented, validated, monitored and overseen by people who can challenge it. AI and Financial Model Risk, a the Global Risks Alliance (GRA) platform, prepares the questions and evidence structures. Validation and approval belong to the institution and its supervisor.
The order things happen in.
- 01QuestionNational, regional or institutional priority
- 02EvidenceSTEM foundation of the platform
- 03BuildFoundry
- 04OrganiseNexus ecosystem
- 05TestUniverse (independent site)
- 06ChapishaRipoti
- 07DecideCompetent authorities and institutions, through NXS-RLS (participants, agreements and handoffs)
- 08RenewCorrections and the next annual agenda
Readiness is not approval.
A readiness file describes a programme’s evidence, risks and governance so that an institution can carry out its own review. GRA does not approve, rate, underwrite, lend or invest, and it recommends no counterparty.
Model and AI rules side by side.
| Regime | Wigo | Key expectation |
|---|---|---|
| SR 26-2 (US, April 2026) | Models at banking organisations; generative and agentic AI stated to be outside scope | Risk-based, tailored model risk management commensurate with model use |
| PRA SS1/23 (UK) | Banks with internal model approval | Five principles: identification, governance, development, validation, mitigants |
| EU AI Act with AI Omnibus | High-risk AI including creditworthiness and life and health insurance pricing | Obligations apply from 2 December 2027 |
| NIST AI RMF | Voluntary, all sectors | Govern, map, measure, manage |
How the process runs.
Process explorer · Process
Resilience and finance readiness: Model risk lifecycle
10 steps·3 stages
Mchakato
Resilience and finance readiness: Model risk lifecycle
US bank supervisors revised model risk guidance on April 17, 2026, with Fed SR 26-2 superseding SR 11-7; it sets no enforceable standards.
01 Inventory and development
- 02Model inventory and materialityDataset · Internal auditAn effective inventory shows model risk for single models and in aggregate. Model purpose with model exposure sets materiality.
- 04Development and testingStep · Internal auditDevelopment starts from a clear purpose aligned with business use and policy. Testing checks whether the model performs as intended.
- 05DocumentationRecord · Internal auditAdequate documentation supports continuity of operations, the tracking of recommendations, responses and exceptions, and remediation.
02 Validation and challenge
- 03Effective challengeComponent · Internal auditObjective experts critically evaluate model risk and drive change across the lifecycle. They need expertise, independence and standing.
- 06Conceptual soundnessStep · Internal auditValidation of conceptual soundness assesses and documents model design, construction and developmental testing.
- 07Outcomes analysisStep · Internal auditOutcomes analysis compares model outputs with real-world outcomes. It can include back-testing or outlier analysis.
03 Governance, use and monitoring
- 01Policies, roles and responsibilitiesComponent · Internal auditPolicies set expectations and a framework for assessing model risk. Roles are clear, with attention to conflicts of interest.
- 08Decision to use, with limits where neededDecision · Internal auditValidation generally comes before first use. Urgent earlier use calls for attention to limits, informed stakeholders and controls.
- 09Ongoing monitoringModel · Internal auditMonitoring checks whether the model performs as expected as products, exposures, clients, data or markets change.
- 10Adjust, recalibrate or redevelopStep · Internal auditWhen performance deviates meaningfully from expectations, the organisation considers adjustment, recalibration or redevelopment.
Outline of an external standard or method, from its public sources. It describes what the framework says, not a record of any run. It does not show that Nexus or any partner performs, certifies or approves a step. A gate is a step of the method, never a Nexus approval. The guidance is expected to be most relevant to banking organisations with over $30 billion in total assets.
Generative AI sits between regimes.
With SR 26-2 placing generative and agentic AI outside its scope, US banks govern those systems through broader risk management and third-party expectations while supervisors develop specific guidance. GRA reports this as a gap in the record, not as a rule. Topic depth at generative AI in finance; oversight at the Global Risks Forum (GRF) AI oversight and accountability.
Who sees a readiness file.
A file is released only at the programme owner’s written direction, to a party the owner chooses, with no fee tied to any transaction.
Where this page’s references lead.
The bodies and documents this page is read against. Status and editions change; each link goes to the issuing body.
- US interagency revised guidance on model risk management, SR 26-2 (April 2026), superseding SR 11-7
- PRA SS1/23 Model risk management principles for banks
- EU AI Act (Regulation 2024/1689) as amended by the AI Omnibus, Regulation 2026/1744 (credit-scoring high-risk duties from 2 December 2027)
- NIST AI Risk Management Framework 1.0
Pages in Resilience and finance readiness.
Questions and answers.
Does a readiness file mean finance is available?
Bring an exposure or capital question to the Alliance.
Information and enquiries are open today. An enquiry receives a reference and names the next step; applications, payments, compute allocations and registrations open only when their team, terms and review capacity exist.
Legal documents.
Maps, data and figures on this site are shown for context and general information. They are not advice, forecasts or warnings; the limits of the information are set out in the Disclaimers and Data Notices.