SOC 2 Control Implementation Guide

Confidentiality / Privacy

Confidentiality Classification for SOC 2

Data classification and handling rules define Confidential, Restricted, Customer Confidential, Internal, and Public information treatment and support identification of confidential information.

Use this guide to put the control into operation, decide what records to retain, and check that an auditor can trace the evidence back to the work your team performed.

Maintained by GreenHat Security · Reviewed August 21, 2026

What this control should accomplish

Personnel can consistently identify confidential information and apply handling rules appropriate to categories such as Restricted, Customer Confidential, Confidential, Internal, and Public data.

First SOC 2 program

A credible starting point

Define a small, understandable classification scheme using categories that fit the company, with examples drawn from customer content, credentials, product designs, contracts, personnel records, support material, and public information. Pair each category with concrete handling behavior.

As the company scales

Make it repeatable

Manage classifications in the data and asset inventories, apply labels in collaboration and data platforms, and connect them to access, encryption, sharing, monitoring, retention, and disposal. Use discovery and data-loss controls to find material that is unclassified or handled inconsistently.

How to implement Confidentiality Classification

  1. 1

    Define the categories

    Describe each classification with decision rules, business examples, owner, and how to resolve information that could fit more than one category.

    You should end up with: An approved classification standard with examples and decision path.

  2. 2

    Create the handling rules

    For every category, state approved storage, access, sharing, transfer, encryption, logging, retention, disposal, and non-production use.

    You should end up with: A concise handling matrix that personnel and system owners can apply.

  3. 3

    Classify important information

    Have owners classify customer data stores, credentials, contracts, design records, personnel information, support exports, and other important repositories.

    You should end up with: Inventory entries or repository labels with owner and review date.

  4. 4

    Apply labels in workflows

    Configure available labels and prompts in document, email, storage, ticketing, and data platforms without relying on labels alone for protection.

    You should end up with: Label configuration and sampled labeled records across key systems.

  5. 5

    Teach practical handling

    Give personnel short examples of how to store, share, export, print, discuss, and dispose of each category and where to ask about ambiguity.

    You should end up with: Role-relevant communication or training record with the current standard version.

  6. 6

    Review inconsistent handling

    Periodically sample repositories and sharing activity, investigate unclassified or misclassified information, and track exceptions to resolution.

    You should end up with: A review report with findings, owners, and corrective actions.

Evidence to keep, and what it should prove

Build the evidence set in three layers: what defines the control, who approved or reviewed it, and what proves it operated. Collect operating records when the work happens so they remain dated, attributable, correctly scoped, and traceable to the underlying activity.

Before sharing, remove unrelated personal or customer data, never expose passwords, tokens, or secret values, preserve enough source context to authenticate the record, and use the secure exchange approved for the engagement.

Policy / design artifacts

Documents that define the control, its scope, ownership, and expected way of working.

Classification standard

  • Confirm what the record proves

    Shows the approved definitions, decision rules, examples, ownership, and handling expectations for Restricted, Customer Confidential, Confidential, Internal, Public, or other adopted categories.

  • Include this context

    classification levels

  • Include this context

    decision criteria and examples

  • Include this context

    required handling by level

  • Include this context

    exception or escalation path

  • Include this context

    owner and approver

  • Include this context

    effective and review dates

Weak evidence to avoid

A glossary of labels with no decision rule, company examples, required handling, owner, or effective date.

data inventory

  • Confirm what the record proves

    Shows which repositories and data sets contain confidential information, their assigned category, owner, location, purpose, and lifecycle context.

  • Include this context

    data set and repository identifier

  • Include this context

    data categories

  • Include this context

    classification

  • Include this context

    system location

  • Include this context

    owner and purpose

  • Include this context

    review date

Weak evidence to avoid

A systems list that does not identify confidential data sets, classification, owner, location, or operational copies.

customer data map

  • Confirm what the record proves

    Shows where customer-confidential information enters, is stored, copied, processed, shared, backed up, and leaves, so its classification follows the full path.

  • Include this context

    customer-data category

  • Include this context

    collection source

  • Include this context

    repositories and copies

  • Include this context

    processors and recipients

  • Include this context

    flow direction

  • Include this context

    owner and review date

Weak evidence to avoid

A map ending at the production database that omits logs, backups, analytics, support attachments, exports, and vendors.

handling quick reference

  • Confirm what the record proves

    Shows personnel the concrete storage, sharing, transfer, access, retention, and disposal behavior expected for each classification in an usable form.

  • Include this context

    classification level

  • Include this context

    approved storage and access

  • Include this context

    sharing and transfer rules

  • Include this context

    retention and disposal action

  • Include this context

    help or exception contact

  • Include this context

    version date

Weak evidence to avoid

A poster saying protect confidential data without explaining approved storage, sharing, transfer, retention, or escalation behavior.

Which records should you prepare for the audit?

Type 1

Evidence at the as-of date

The classification standard, current confidential-data inventory and customer-data map, live handling reference, and all unclassified, disputed, or exception items on the examination date.

Type 2

Evidence across the review period

Every new or materially changed information repository, customer-data path, support or collaboration location, export process, or third-party transfer during the review period that required classification, plus every scheduled owner review, reclassification, and handling exception decided during the period.

Completeness check

Reconcile current cloud, SaaS, collaboration, support, data-store, and discovery inventories to classified data records and customer flows; compare period repository and transfer changes with owner decisions and investigate every unclassified location, lost label, or handling exception.

Build the record set from

  • data catalog and asset inventory
  • cloud and SaaS repository inventory
  • customer data-flow repository
  • document labeling and data discovery
  • production change system
  • classification exception queue

Keep these fields for each record

  • data set or repository identifier
  • creation, change, or review trigger
  • data categories and classification
  • owner and location
  • handling requirements
  • decision and effective date
  • exception or remediation status

How an auditor may test this control

Use this checklist to prepare for procedures an auditor may perform. The exact steps and sample selection depend on your engagement scope and the service auditor's professional judgment.

  • Confirm the intended control outcome

    Determine whether the control is designed to achieve this result: Personnel can consistently identify confidential information and apply handling rules appropriate to categories such as Restricted, Customer Confidential, Confidential, Internal, and Public data.

  • Confirm ownership and operating cadence

    Compare the documented owner with the intended role (Privacy Owner / CISO / Data Owners), then compare dated records with the stated cadence: Ongoing; periodic review by privacy/data owner.

  • Establish the complete audit record set

    Reconcile current cloud, SaaS, collaboration, support, data-store, and discovery inventories to classified data records and customer flows; compare period repository and transfer changes with owner decisions and investigate every unclassified location, lost label, or handling exception.

  • Prepare the as-of-date evidence for a Type 1 engagement

    The classification standard, current confidential-data inventory and customer-data map, live handling reference, and all unclassified, disputed, or exception items on the examination date.

  • Prepare period evidence for a Type 2 engagement

    Every new or materially changed information repository, customer-data path, support or collaboration location, export process, or third-party transfer during the review period that required classification, plus every scheduled owner review, reclassification, and handling exception decided during the period.

  • Inspect the policy / design artifacts

    Documents that define the control, its scope, ownership, and expected way of working.

    • Inspect Classification standard

      For each selected record, confirm it demonstrates Shows the approved definitions, decision rules, examples, ownership, and handling expectations for Restricted, Customer Confidential, Confidential, Internal, Public, or other adopted categories.

      • classification levels
      • decision criteria and examples
      • required handling by level
      • exception or escalation path
      • owner and approver
      • effective and review dates
    • Inspect data inventory

      For each selected record, confirm it demonstrates Shows which repositories and data sets contain confidential information, their assigned category, owner, location, purpose, and lifecycle context.

      • data set and repository identifier
      • data categories
      • classification
      • system location
      • owner and purpose
      • review date
    • Inspect customer data map

      For each selected record, confirm it demonstrates Shows where customer-confidential information enters, is stored, copied, processed, shared, backed up, and leaves, so its classification follows the full path.

      • customer-data category
      • collection source
      • repositories and copies
      • processors and recipients
      • flow direction
      • owner and review date
    • Inspect handling quick reference

      For each selected record, confirm it demonstrates Shows personnel the concrete storage, sharing, transfer, access, retention, and disposal behavior expected for each classification in an usable form.

      • classification level
      • approved storage and access
      • sharing and transfer rules
      • retention and disposal action
      • help or exception contact
      • version date
  • Trace the control from design to operation

    Use the categories that apply to this control: connect any policy or design artifact to its approval or review record, then trace a selected operating record through execution, result, and any exception or remediation.

Common implementation and evidence gaps

  • Classification terms differ across policy, product, contracts, and technical systems.
  • Nearly everything is labeled confidential, so personnel cannot distinguish higher-risk handling.
  • Labels are defined without access, transfer, retention, or disposal consequences.
  • Exports, screenshots, email attachments, support tickets, and collaboration copies lose the source classification.
  • Customer-specific confidentiality commitments are absent from owner decisions.
  • Automated labels are trusted without review of false positives, missed repositories, or unstructured content.

Before you call this control ready

  • Can personnel classify common examples and explain the handling difference between levels?
  • Do sampled repositories show classifications approved by an accountable data or business owner?
  • Are access, sharing, encryption, retention, and disposal consistent with the assigned level?
  • Do exports, support records, email, and collaboration systems preserve or compensate for classification?
  • Are discovered exceptions and ambiguous cases assigned and resolved?

Trust Services Criteria references

These identifiers help you navigate related Trust Services Criteria. They do not reproduce the criteria or prove that this control fully addresses them in your environment.

  • CC2.1
  • C1.1

Confirm final scope, mappings, and testing expectations with your service auditor. SOC 2® is an AICPA trademark; GreenHat Security is not affiliated with or endorsed by AICPA.