Standards, Controls & Policy Execution
Author, maintain, and version the bank's AI standards library: model selection and approval, human oversight tiers, prompt/response logging and retention, agentic guardrails, and acceptable-use requirements.
Build and maintain the obligation-to-control-to-evidence mapping against applicable regulatory guidance (SR 26-2, NYDFS, and evolving supervisory expectations).
Operate the AI policy exception and TRM deviation processes: draft time-bound deviation requests, define compensating controls, track expiries and remediation paths to closure.
Support annual policy refresh cycles and translate policy principles into testable, enforceable requirements in partnership with policy owners.
Platform-Enforced Governance & Evidence
Partner with the AI platform team to ensure governance requirements are enforced mechanically at the enterprise AI gateway (entitlements, model allowlists, quotas, audit logging) and that enforcement produces native evidence.
Define evidence requirements and validate audit artifacts as queryable, export-ready, and sufficient for internal audit, second-line review, and examiner requests.
Design and operate continuous-monitoring practices: monitoring boundaries, event-based review triggers, and evidence expectations for AI systems that change frequently.
Maintain the AI use case and agent registry as the system of record, integrated with platform onboarding and entitlement workflows.
Provide governance ownership of the AI use-case intake framework in partnership with the AI Use Case Intake & Value Lead.
Risk Partnership & Regulatory Readiness
Serve as the primary working-level interface to Technology Risk Management, Model Risk Management, Compliance, and Internal Audit for AI governance matters.
Assemble examiner-readiness packages: control mappings, evidence samples from the live platform, and documented rationale for the standards the bank has defined.
Monitor the regulatory and industry landscape (agency guidance, FSB/trade-group developments) and assess impact to standards and controls.
Support AI Control Group and AI Ethics Working Group activities with documented assessments and control recommendations.
