Federal Modernization Workbook · Book 20
Book 20 — Information Protection & Microsoft Purview
Protection That Travels With the Data
Why protection has to travel with the data: classes and owners, sensitivity labels and defaults, container labels, Rights Management and recovery, message encryption, SharePoint and OneDrive, DLP, auto-labeling, audit and insider risk, AI data security, and the operations and evidence that keep it working
This book is new material, written for the Zero Trust Assessment closure update of November 2026, and every page is a draft awaiting the editor’s review. It isn’t part of Release 2.0, the baseline edition (Status labels). Every external claim was checked against a primary source on November 11, 2026, Microsoft Learn first for product behavior, licensing and government-cloud availability, and each chapter keeps its verification log in the repository. Purview features differ a great deal between the commercial cloud, GCC, GCC High and DoD, and change often: each chapter states availability per feature from Learn as read on that date, and says “not yet specified” where Learn is silent. The commands are taken from Learn and haven’t been run against a production tenant; run them against a test tenant first.
b20-information-protection-bundle.docx: the whole book as one Word document, regenerated on every site deploy.
An agency that has hardened identity, devices and the network still lets a legitimate user on a compliant device send a file to an approved destination, whatever the file holds. Nothing upstream said what the payload was. Book 20 is the workbook’s home for the Zero Trust data pillar: deciding what the agency’s data classes are and who owns them, turning them into sensitivity labels that travel with the content, encrypting what has to stay protected outside the tenant and keeping a tested way back into it, stopping sensitive data from leaving through DLP, finding what users didn’t label, watching it through audit and insider risk, and governing what AI assistants can reach. It is engineered with Microsoft Purview, and every procedure follows the workbook’s implementation conventions: the five control states, the How-section template, authority labels, the evidence checklist, change safety and the Zero Trust mapping block (Book 00, Appendix D).
Book 20 links where another book already owns the subject. The federal CUI framework and its mapping to labels, the label taxonomy as code, the first DLP baseline and the three meanings of “customer-managed key” are in Book 12, Chapter 09; the data platform and the SQL Server estate are Book 06; telemetry, detection and the evidence contract are Book 12; authorization, privacy and the control crosswalk are Book 14.
Why, what, who, how
| Question | Chapters | |
|---|---|---|
| Why | Why does protection have to travel with the data, and what has to be decided first? | 01 |
| What | The reference architecture and the sensitivity label design | 02 |
| How | Publishing and defaults, containers, Rights Management, message encryption, SharePoint and OneDrive, DLP, auto-labeling, audit and insider risk, and AI data security | 03, 04, 05, 06, 07, 08, 09, 10, 11 |
| Who | Who operates it, and what an assessor can inspect | 12, Appendix A |
The book’s required artifacts
| Artifact | Where |
|---|---|
| Reference architecture diagram, with trust boundaries and control planes | Chapter 02 |
| Decision record template, with license, cloud availability, prerequisite and compensating-control fields | Chapter 01 |
| Implementation checklist, with pilot, enforcement, rollback, exception and validation steps | Chapter 12 |
| RACI: policy owner, platform owner, operator, security monitor, help desk and assessor | Chapter 12 |
| Evidence catalog: configuration, test, operational, exception and reassessment artifacts | Chapter 12 |
| Metrics of protected coverage and control effectiveness | Chapter 12 |
Reading paths
- CIOs, CISOs and data owners (20 minutes): Chapters 01 and 12, and the reference architecture in Chapter 02.
- Purview and Microsoft 365 engineers: Chapters 02 to 09, in order.
- Records officers and privacy officials: Chapters 01, 05 and 10, with Book 14, Chapter 08.
- Security operations: Chapters 08, 10 and 12, with Book 12.
- AI program owners: Chapter 11, with Book 17.
- ISSOs and assessors: Chapter 12 and Appendix A, with Book 11, Chapter 10 for the closure chain.
Where this book sits in the workbook
This is Book 20 of the Federal Modernization Workbook. It owns the data pillar of the agency’s Zero Trust program (Book 11). The groups that labels grant rights to come from the identity lifecycle in Book 03; Conditional Access, PIM and emergency access are Book 04; managed and onboarded devices are Book 05; drift detection is Book 13; the help desk is Book 15; user training on labels and encrypted mail is Book 16; and identities for AI agents are Book 17.
Chapters
| Title | Description |
|---|---|
| Chapter 01 — Data Inventory and Classification | Why protection has to travel with the data, and the decisions that come before any label: the agency’s authoritative data classes and the handling outcome each one needs, who owns each class, the federal overlay decisions (CUI categories, Privacy Act records, federal records, the FIPS 199 impact level) that set the classes, how to inventory where sensitive data already sits with Microsoft Purview’s data classification tools, licensing and government-cloud availability as stated dependencies, and the decision record template every later chapter uses. |
| Chapter 02 — Sensitivity Label Architecture | The reference architecture for information protection, with its trust boundaries and control planes, and the label design built on it: a small, usable taxonomy taken from the class register; label scopes for files, email, meetings, and groups and sites; label priority and label groups (the modern label scheme); encryption with administrator-defined rights to mail-enabled groups; content markings and their limits; publishing policies and their priority; and building the taxonomy in a test tenant first, with each feature’s licensing and government-cloud availability. |
| Chapter 03 — Label Publication and Defaults | How the agency publishes its sensitivity labels and sets defaults it can justify: a pilot label policy before the tenant-wide one, a default label for documents, email, meetings, new groups and sites, and Power BI content, each with its reason and the label it should not be; mandatory labeling and downgrade justification; the exact portal steps and Security & Compliance PowerShell commands, the rings, rollback and exclusions; the telemetry and evidence that show labels are being applied; and each feature’s licensing and government-cloud availability, including the SharePoint library default label that is still in development in GCC, GCC High and DoD. |
| Chapter 04 — Teams, Groups and SharePoint Containers | How container sensitivity labels govern the places content lives: privacy, guest access, external sharing, access from unmanaged devices, authentication contexts, private-team discovery and shared channels, the default sharing link and members’ sharing rights; who may create new teams, groups and sites; labeling existing sites in bulk; the Conditional Access dependency for unmanaged devices; the document-to-site mismatch signal; and the exact commands, rings, rollback, exclusions, telemetry, evidence and government-cloud availability for each. |
| Chapter 05 — Rights Management and Recovery | How the agency turns on the encryption service behind every protecting sensitivity label and every encrypted message, and how it keeps the ability to read what it encrypted: confirming and activating the Azure Rights Management service, onboarding controls for a phased start, the government-cloud connection and DNS steps, licensing as a stated dependency, the super user feature held off by default and enabled just in time through Privileged Identity Management, and a decryption-continuity test that proves a holder can open, a non-holder can’t and a recovery operator can, with the evidence each step leaves. |
| Chapter 06 — Exchange Message Encryption | How the agency encrypts the mail that leaves it without asking every sender to remember: when to use a sensitivity label and when a mail flow rule, the approved message scenarios, verifying Microsoft Purview Message Encryption with Test-IRMConfiguration, branding, one-time passcodes and social sign-in, the Encrypt button in Outlook on the web, Advanced Message Encryption’s custom templates, expiry and revocation, mail flow rules staged from audit to enforcement by ring, rollback, exclusions, telemetry, and the encrypted test and transport-rule evidence that closure-map row CM-13 asks for, with each feature’s availability in GCC, GCC High and DoD. |
| Chapter 07 — SharePoint and OneDrive | How the agency makes labeled and encrypted files work where most of them live: why SharePoint and OneDrive can’t search, scan, co-author or apply DLP to encrypted files until label integration is on; turning it on, with PDF support, and the encryption settings it can’t process; inventorying legacy SharePoint IRM libraries and moving their protection to sensitivity labels, with the government-cloud gap in default library labels; turning on co-authoring for encrypted files only after the labeling-metadata change is safe for every app that reads labels; and the rollback, telemetry, evidence and ownership for each, with feature availability in GCC, GCC High and DoD. |
| Chapter 08 — DLP and Endpoint Controls | How to design Microsoft Purview Data Loss Prevention from data outcomes rather than from templates: the outcome statement, the surface and the action for each one; simulation mode as the pilot; policy tips, overrides and business justification; false positives and tuning; exceptions that don’t become holes; endpoint DLP on onboarded Windows devices; DLP incident handling; and the How sections, with government-cloud availability, licensing, change safety, rollback, telemetry and evidence. |
| Chapter 09 — Auto-Labeling and Sensitive Information Types | How the agency detects sensitive content it never labeled: the classifier ladder from built-in sensitive information types to custom types, exact data match and trainable classifiers; confidence, count and proximity; custom types and rule packages as code; exact data match with the two-computer hash and upload; client-side and service-side auto-labeling; and the How section that takes a service-side auto-labeling policy from simulation to rollout, with government-cloud availability, licensing, limits, rollback, telemetry and evidence. |
| Chapter 10 — Audit, Insider Risk and Communication Compliance | The tools that watch activity and behavior around the agency’s data, and the governance they need first: Purview Audit Standard and Premium, retention and routing; Insider Risk Management and Communication Compliance under a charter that settles privacy, legal and labor-relations questions, role separation, pseudonymization, triage, retention and evidence; and the How sections that turn on, retain and test Purview Audit with searchable synthetic events, and stand up Insider Risk Management and Communication Compliance, with government-cloud availability, licensing, change safety, rollback, telemetry and evidence. |
| Chapter 11 — AI Data Security | How the agency monitors and governs sensitive data in approved and unapproved AI workflows: how Microsoft 365 Copilot honors permissions, sensitivity labels and the EXTRACT and VIEW usage rights; Data Security Posture Management for AI and the current Data Security Posture Management, with their prerequisites, one-click policies, activity explorer and data risk assessments for oversharing; DLP for Microsoft 365 Copilot to keep labeled content out of Copilot processing and sensitive prompts out of web grounding; third-party AI sites through endpoint and browser DLP; the government-cloud availability of each feature; and the control states, procedure, evidence and day-two ownership. |
| Chapter 12 — Operations and Evidence | Who runs information protection after it is built, and how the agency proves it keeps working: change control for labels, policies and keys with a dated configuration export; drift detection against that baseline; incident response for data events; reassessment on the closure chain; the assessor artifacts; and four of the book’s required artifacts: the implementation checklist for every control in Chapters 03 to 11, the RACI separating policy owner, platform owner, operator, security monitor, help desk and assessor, the evidence catalog, and metrics that measure protected coverage and control effectiveness rather than configuration presence. |
| Appendix A — Requirements Mapping: Controls, KSIs and Chapters | A compliance-enabling map, not a compliance claim: the NIST SP 800-53 Rev. 5 controls Book 20’s chapters support, with what each control asks, the chapters that address it, the book that owns the rest and the evidence an assessor can expect; CISA’s Zero Trust Maturity Model Data pillar functions; OMB M-22-09’s data actions and the binding instruments beside them; FedRAMP’s 2026 Key Security Indicators used as an evidence vocabulary; the closure-map rows the book carries; and the sources. |
Version and revision history
This book has not been published in an edition yet. Each edition is kept, with its Word editions, decks and videos, on the Archive page. Every change so far is in the commit history.