Discover the latest trends and network with industry leaders and explore innovative solutions.  View Events

REGULATIONS & FRAMEWORKS

One set of controls. you answer to.

Most enterprises answer to several AI rules at once. A3R builds one set of controls and maps it to each framework, so you aren’t running a separate program for every regulation.

How the mapping works

A3R’s controls are organized around what runtime-enforced governance looks like: visibility by default, lineage by design, policy enforced at the point of interaction, and accountability mapped to action. Each control produces its evidence once, and that evidence answers the matching requirement in every framework that applies to you.

How our controls map

A3R control areaNIST AI RMFISO/IEC 42001EU AI Act (high-risk)Sector rules
AI inventory and risk tieringGovern, MapClause 6 (planning), Annex AArt. 6 classification, Art. 9 risk managementSR 26-2 (banking), NAIC model bulletin (insurance)
Data lineage and qualityMap, MeasureAnnex A data controlsArt. 10 data governanceHIPAA (healthcare)
Policy enforced at runtimeManageClause 8 (operation)Art. 9 risk management, Art. 15 accuracy and robustnessABA Formal Opinion 512 (law)
Human oversight and accountabilityGovernClause 5 (leadership)Art. 14 human oversightAll sectors
Testing and monitoringMeasureClause 9 (performance evaluation)Art. 15, post-market monitoringSR 26-2, NAIC model bulletin
Records and evidenceGovern, ManageClause 7.5 (documented information)Art. 11 technical documentation, Art. 12 record-keepingAll sectors
Vendor and third-party AIGovern, MapAnnex A supplier controlsDeployer obligationsSR 26-2, HIPAA business associates

This mapping is a summary for orientation. The assessment includes a requirement-by-requirement gap report for the frameworks that apply to you.

What changed in 2026

EU AI Act

High-risk obligations now apply from December 2, 2027 for stand-alone systems and August 2, 2028 for systems embedded in regulated products. Transparency obligations were not deferred.

US banking

In April 2026, the Federal Reserve, FDIC, and OCC replaced SR 11-7 with new model risk management guidance (SR 26-2 / OCC Bulletin 2026-13) that takes a more risk-based approach.

US states

State AI laws are changing quickly. Colorado’s law did not take effect as planned, and federal efforts to preempt state laws are being contested. Until courts or Congress decide, state AI laws remain enforceable.

Last reviewed: October 2026.

By industry

Model risk management under SR 26-2, fair lending, and exam evidence.

The NAIC model bulletin adopted by many states, and testing for unfair discrimination.

HIPAA, and clinical summaries every doctor can check against the record.

Law

ABA Formal Opinion 512 and client confidentiality.

Safety, separation from control systems, and documentation for regulators and rate cases.

This page is a summary, not legal advice. Work with your counsel on how each rule applies to your organization.

Start with the assessment.

In about four weeks, you’ll know what AI you’re running, where you stand on the maturity curve, and what to fix first.

Subscribe to our Newsletter
Stay up-to-date with the latest AI and data insights from a3r.

© 2026 a3r. All rights reserved.