New, with Accorian: a real-time AI governance framework for control drift in enterprise AI.Read the framework
AI Governance · SR 11-7
SR 11-7 for AI and model risk
For banks and capital-markets teams: AI behavior turned into the inventory, validation, monitoring, and governance evidence model-risk teams already use. The bank-grade discipline US examiners apply, mapped to live AI.
GuidanceSR 11-7
IssuerFed / OCC
Spine3-part
Model-risk evidence from live AI behavior.
Every claim in the report traces back to source evidence, ownership, and the workflow decision it supports.
Valuefund next
Riskcontain now
Fluencytrain where work changed
What it isType: Supervisory guidanceIssuer: Federal Reserve & OCCDomain: Model-risk management in bankingApplies to AI: When it drives material banking decisions
At a glance
A three-part model-risk spine for AI.
Requirement → Evidence
The guidance names the expectations. We produce the evidence.
SR 11-7 expectationEvidence we produce
Model inventoryA complete inventory with ownership, purpose, use, limits, and materialityAI inventory, model or agent owner, business use, materiality flag, dependency map, approval state
Development & documentationDesign, data, assumptions, limits, and implementation controlsUse-case baseline, data-source record, prompt or model version, control design, test set, signoff
ValidationIndependent assessment of soundness, outcomes, and performanceValidation packet, benchmark results, exception analysis, challenger notes, weakness log, remediation status
Ongoing monitoringTrack performance, monitor change, escalate, report to governanceDrift report, control-health series, incident trace, change log, stale-evidence flag, committee-ready summary
One pipeline, every framework
How the SR 11-7 evidence gets produced.
Trace
production behavior
Classify
sensitivitysource
Evaluate
per-use-case baseline
Map
every framework
Pack
on demand
What teams should remember
Three things SR 11-7 evidence gets right.
01
AI does not escape model risk
If an AI system influences a banking decision, model-risk teams need inventory and monitoring.
02
Validation needs behavioral evidence
Static docs are not enough when outputs change with prompts, data, tools, vendors, or model versions.
03
Governance forums need a summary
We keep raw traces available while producing risk summaries committees can actually review.
SR 11-7 AI, asked plainly
Direct answers.
No. It is banking supervisory guidance for model-risk management. AI can fall under scope depending on use.
Keep the evidence map connected
Pairs cleanly with the rest.
Specialist AI builder, across the board
One builder, across the board.
We take your AI from strategy to outcome, with governance, audit, and evals built into every build. Start with a discovery call, or a quick audit.