The Compliance Atlas ·Directory view
A reference for practitioners · 2026 edition

The Compliance Atlas

11 volumes
4 layers per volume
one editorial language

The directory view of the Atlas — every volume with its full description. Audit work as practiced, not recited from a textbook. Each volume walks one framework through a single editorial structure: the lifecycle in two lanes, the control universe, the evidence and testing, and the crosswalk to the rest.

How to read the Atlas

Each volume is a standalone HTML page. Open one, work through its four layers, and screenshot what's useful. Volumes are siblings — there is no required reading order, but SOX → SOC 2 → ISO 27001 is the canonical sequence for IT auditors.

The Crosswalk in each volume maps that framework's controls to the others, so you can read across the Atlas one control domain at a time once you've internalized one volume.

Eleven volumes total across three parts. Return to the spectrum landing → for the visual map and the reading paths.

Part one

Financial & assurance frameworks

3 volumes
Volume I · Edition 2026.2
I.

Sarbanes–Oxley

An integrated audit, walked side by side

PCAOB AS 2201 top-down approach. ELC, ITGC, ITAC, BPC. Deficiency severity from CD to MW. The 12-month internal × external lifecycle as it actually plays out.

Volume II · 2026.1
II.

SOC 2

Trust Services Criteria, examined

AICPA TSP-100. Five Trust Services Criteria, 33 Common Criteria. Type 1 vs Type 2 mechanics. CUECs, sub-service organizations, the bridge letter, and how a SOC 2 report actually gets used by your customers' auditors.

Volume III · 2026.1
III.

ISO/IEC 27001:2022

An ISMS, certified

Clauses 4–10 + Annex A's 93 controls in 4 themes. Stage 1 + Stage 2 audits, surveillance, recertification. The Statement of Applicability as a living document.

Part two

Cybersecurity & industry-specific frameworks

5 volumes
Volume IV · 2026.1
IV.

NIST CSF 2.0

A profile, not a certification

Six functions (GOVERN added in 2024), 22 categories, 106 subcategories. Current Profile vs Target Profile, Tier 1–4 maturity. Why CSF is read by everyone and audited by no one — until you make it part of a SOC 2+ engagement.

Volume V · 2026.1
V.

HITRUST CSF v11

A unified harmonization

e1, i1, r2 assessment levels. The MyCSF mechanics. Why HITRUST exists where HIPAA, ISO 27001, SOC 2, and NIST already do — and what a HITRUST-certified report actually proves.

Volume VI · 2026.1
VI.

PCI DSS v4.0.1

Cardholder data, scoped tightly

12 requirements, 64 sub-requirements. CDE definition and segmentation testing. QSA-led RoC vs SAQ paths. The customized vs defined approach in v4. Why scope is the only thing that matters.

Volume VII · 2026.1
VII.

HIPAA Security Rule

A safeguard rule, enforced

45 CFR 164 Subparts C & D. Administrative, physical, technical safeguards. OCR audits and breach notification. Required vs addressable specifications — and why "addressable" is misread by most teams.

Volume VIII · 2026.1
VIII.

FedRAMP

An ATO, federal cloud authorized

Built on NIST 800-53 Rev 5. 3PAO-led security assessment, FedRAMP PMO at GSA, JAB or agency authorization paths. High / Moderate / Low baselines (~410 / 325 / 156 controls). The only path to selling cloud services to the federal government.

Part three

AI governance frameworks

3 volumes
Volume IX · 2026.1
IX.

ISO/IEC 42001:2023

An AI management system

The first management-system standard for AI. Clauses 4–10 + Annex A controls + Annex B implementation guidance. Internal vs notified-body lanes — the audit profession arriving at AI without a settled playbook.

Volume X · 2026.1
X.

EU AI Act

Conformity assessed, market surveilled

Risk tiers — prohibited, high-risk, limited-risk, minimal. Annex III high-risk systems. Conformity assessment procedures (Annex VI / VII), notified bodies, the EU AI Office, and national market surveillance authorities. The first real regulatory regime for AI.

Volume XI · 2026.1
XI.

NIST AI RMF

An AI risk profile, voluntary

NIST AI 100-1 (Jan 2023). Four functions — GOVERN, MAP, MEASURE, MANAGE. The 2024 Generative AI Profile (NIST AI 600-1). No native certification, like CSF — but the foundational vocabulary that most AI governance work uses to describe itself.

What lives in every volume

Layer 01 — Lifecycle Flow. A two-lane swimlane showing the audit as practiced. For traditional frameworks: internal audit × external audit. For AI frameworks: internal × notified-body / regulator.

Layer 02 — Control Mindmap. The framework's structure, drilled to test approach.

Layer 03 — Evidence & Testing. Sample sizes, IPE, workpapers, the unglamorous work that makes an audit hold up.

Layer 04 — Crosswalk. This volume's domains mapped to the other ten, with shared evidence column.

Editorial conventions

Internal lane in slate blue. External lane in oxblood. Bridges in gold dashed. Solid box = primary procedure; dashed = follow-on or parallel.

Authoritative paragraph and clause numbers are cited in the small caps margin reference of each layer header.

No fictional companies in this edition. The case-study edition is forthcoming and will use this canon as its testbed.

What this is for

For practitioners walking their second or third audit who need a single reference that doesn't lie about how the work actually goes.

For senior auditors stress-testing their own program against a clean editorial structure.

For junior auditors who want to see the work before they have to do it.

For founders, CISOs, and engineers asking why their compliance team keeps asking for the same thing seven different ways.