Glossary
The language of AI compliance.
Plain-language definitions of the terms a financial-services compliance or security team meets when governing AI — each linked to the controls and frameworks that address it.
Examination packageThe bundle of evidence a firm hands to a regulator or examiner.FINRA Regulatory Notice 24-09FINRA's 2024 reminder that existing rules apply to generative AI.MNPI (Material Non-Public Information)Information that could move a security's price and hasn't been made public.Model riskThe risk of loss from decisions based on incorrect or misused models.NYDFS 23 NYCRR 500New York's cybersecurity regulation for financial-services entities.PII (Personally Identifiable Information)Data that can identify an individual, subject to privacy and security controls.SR 11-7The Fed/OCC supervisory guidance on model risk management.Tamper-evident audit chainA hash-linked sequence of records where any alteration breaks the chain.WORM (Write Once, Read Many)Storage that can be written once and never altered — the basis of tamper-evident records.
See it in practice
From definitions to evidence.
See how these controls produce an examination-ready audit trail on your own traffic.