Information Classification Scheme

To meet the information classification scheme requirement in VDA ISA 2.1.1, you must define and implement a practical, organization-wide classification model (at least Public, Internal, Confidential, Strictly Confidential) that matches OEM partner expectations and common industry practice. Then you must operationalize it through labeling, handling rules, access controls, training, and evidence that it is consistently applied. (VDA ISA Catalog v6.0)

Key takeaways:

  • Your scheme must align with OEM partner classification requirements, not just internal preference. (VDA ISA Catalog v6.0)
  • A classification policy alone is insufficient; you need handling rules, ownership, tooling support, and proof of adoption. (VDA ISA Catalog v6.0)
  • Auditors will test for consistency across systems, teams, and third parties, especially for high-sensitivity engineering and customer data. (VDA ISA Catalog v6.0)

An information classification scheme is the backbone of how you control information risk across engineering, manufacturing, corporate IT, and third-party collaboration. Under VDA ISA 2.1.1, the bar is straightforward: define and implement a scheme consistent with OEM partner requirements and industry standards. (VDA ISA Catalog v6.0) The operational challenge is that “classification” quickly turns into a messy debate about labels, exceptions, and how to handle mixed datasets like CAD files with embedded supplier data or test results exchanged over collaboration platforms.

A workable approach starts with a small set of levels and makes the handling rules unambiguous. People need to know what to do differently for “Confidential” versus “Strictly Confidential” in real workflows: where the data can be stored, how it can be shared, whether it can be emailed, what encryption is required, who approves external sharing, and how long it is retained. Your assessors will look for evidence that these rules are implemented in systems and behaviors, not just written down.

This page translates the requirement into an execution playbook a CCO or GRC lead can run: define the scheme, map it to OEM expectations, deploy it into tools and processes, and retain the artifacts that make audits predictable. (VDA ISA Catalog v6.0)

Regulatory text

Requirement (VDA ISA 2.1.1): “Define and implement an information classification scheme consistent with OEM partner requirements and industry standards.” (VDA ISA Catalog v6.0)

Operator meaning: you need (1) a defined classification model and (2) implementation proof that the model is used across relevant information types and workflows. “Consistent with OEM partner requirements” means you must explicitly map your levels and handling rules to what your OEM customers expect in contracts, security exhibits, data exchange requirements, or collaboration portal rules. (VDA ISA Catalog v6.0)

Minimum expected levels (practical baseline reflected in the provided summary):

  • Public
  • Internal
  • Confidential
  • Strictly Confidential (VDA ISA Catalog v6.0)

Plain-English interpretation (what an assessor is really checking)

An assessor will test whether your organization can reliably answer four questions for any meaningful dataset (drawings, prototypes, specifications, test data, incident reports, pricing, supplier terms, HR data):

  1. What is this information’s classification? (clear level definitions)
  2. Who can access it and why? (access decisions tied to classification)
  3. How can it be stored, transmitted, and shared? (handling rules)
  4. Can you prove it is applied in day-to-day work? (evidence across systems and third parties)

If your teams “decide case-by-case” without consistent rules, you will struggle to show implementation even if the policy exists. (VDA ISA Catalog v6.0)

Who it applies to

Entity types: automotive suppliers and OEMs pursuing TISAX-aligned assurance against VDA ISA requirements. (VDA ISA Catalog v6.0)

Operational contexts where this requirement bites hardest:

  • Engineering and R&D collaboration with OEMs and their delegated third parties (design data, CAD, software, requirements documents)
  • Manufacturing and quality (process parameters, inspection results, defect data)
  • Procurement and supply chain (RFQs, pricing, supplier contracts, tooling arrangements)
  • Corporate functions (finance, HR, legal) where regulated or sensitive information is handled
  • Cross-boundary sharing: email, file transfer, portals, collaboration suites, ticketing systems, and managed service providers (VDA ISA Catalog v6.0)

What you actually need to do (step-by-step)

Use this sequence to get to “defined and implemented” without getting stuck in taxonomy debates.

1) Inventory the information that matters (scope the scheme)

Create a short list of “information domains” that reflect how your business runs. Keep it operational, not academic:

  • Product/engineering data (CAD, BOMs, requirements, source code)
  • Customer/OEM data exchanges (portal exports, shared folders, statements of work)
  • Manufacturing/quality data (control plans, test results)
  • Commercial data (pricing, bids)
  • Corporate sensitive data (HR, financials, legal)

Deliverable: an Information Domain Register with domain owner, typical repositories, and primary third parties involved.

2) Define the classification levels with decision rules

Write definitions people can apply quickly. Include:

  • Definition (what qualifies)
  • Impact rationale (what could go wrong if disclosed/altered)
  • Examples (3–5 concrete examples per level)
  • Default rule (what to pick when unsure)
  • Downgrade/upgrade conditions (who can change classification and why)

Minimum levels should include Public/Internal/Confidential/Strictly Confidential. (VDA ISA Catalog v6.0)

Deliverable: Information Classification Standard (1–3 pages) plus a one-page “quick guide.”

3) Map your levels to OEM partner requirements (do not skip)

Build a mapping table for each key OEM partner requirement set you must follow (contracts, security exhibits, data handling rules). The mapping should show:

  • OEM label → your label
  • Required controls (e.g., encryption, storage location constraints, sharing restrictions) → your handling rules
  • Any stricter OEM rule that overrides your default

Deliverable: OEM Classification Mapping Matrix reviewed by Legal/Commercial plus InfoSec/GRC. (VDA ISA Catalog v6.0)

4) Define handling rules that change behavior

For each classification level, specify enforceable handling controls:

  • Storage: approved repositories; restrictions on personal drives or unmanaged cloud storage
  • Access: role-based access expectations; approval requirements for “Strictly Confidential”
  • Transmission: encryption requirements; approved channels for external transfer
  • Sharing with third parties: when NDAs/DPAs are required; who approves; how access is revoked
  • Labeling: file headers/footers, metadata tags, folder naming conventions
  • Retention and disposal: tie into your retention schedule and secure deletion process

Deliverable: Information Handling Rules table (one page) that you can paste into procedures and onboarding.

5) Implement in tools and workflows (prove adoption)

Pick implementation points that leave evidence:

  • Document templates (classification header/footer)
  • Collaboration platforms (default folder labels; restricted groups)
  • Ticketing/engineering change workflows (required classification field)
  • Data room / file transfer process (approval gates for “Strictly Confidential”)
  • Joiner/mover/leaver access workflows tied to classification and domain ownership

Deliverable: System configuration screenshots, workflow configs, and template samples that show the scheme is embedded.

6) Assign ownership and exception handling

Define:

  • Information owners per domain (accountable for classification decisions)
  • Custodians (IT/system owners responsible for technical enforcement)
  • Exception process (who can approve exceptions, for how long, and required compensating controls)

Deliverable: RACI and Exception Log (with approvals and expiry).

7) Train and test

Train role-specifically:

  • Engineers: how to classify CAD/software artifacts; external collaboration rules
  • Procurement: RFQ and pricing confidentiality
  • IT/admins: how to enforce labeling and access groups
  • Third-party managers: sharing rules and contract alignment

Then test with spot checks: select samples from each domain and verify classification + handling matches the standard.

Deliverable: training completion records, quiz/attestation, and classification spot-check results with corrective actions.

Required evidence and artifacts to retain

Auditors want implementation proof. Retain:

  • Information Classification Policy/Standard (current and prior versions)
  • Handling rules matrix by classification level
  • OEM partner mapping matrix and sign-off (VDA ISA Catalog v6.0)
  • Information domain register and named owners
  • System configurations (screenshots, settings exports) showing labeling/access enforcement
  • Templates with classification markings
  • Training materials and completion records
  • Exception log with approvals and expiry
  • Spot-check/audit reports and remediation tickets

Tip: store artifacts in a single audit-ready folder keyed to VDA ISA 2.1.1 so you can answer requests fast.

Common exam/audit questions and hangups

Expect these:

  • “Show me how an engineer decides between Confidential and Strictly Confidential.”
  • “Where is the OEM mapping documented and approved?” (VDA ISA Catalog v6.0)
  • “Demonstrate how classification is applied in your main repositories (PLM, SharePoint, ticketing).”
  • “How do you control external sharing to third parties for Strictly Confidential data?”
  • “Show exceptions and prove they expire and are reviewed.”
  • “Give examples of misclassifications found and how you corrected them.”

Hangups that trigger follow-ups:

  • Definitions exist, but staff cannot apply them consistently.
  • Handling rules are vague (“use secure methods”) instead of channel-specific.
  • “Strictly Confidential” exists on paper but has no stricter handling than “Confidential.”
  • OEM requirements are referenced but not mapped and operationalized. (VDA ISA Catalog v6.0)

Frequent implementation mistakes (and how to avoid them)

  1. Too many classes or sub-classes.
    Fix: keep four levels, add domain-specific examples instead of new labels.

  2. No default classification rule.
    Fix: set a default (commonly “Internal”) and require escalation for external sharing.

  3. Labeling without handling rules.
    Fix: write “if/then” rules per channel (email, file transfer, portal, removable media).

  4. Relying on training only.
    Fix: embed classification into tooling (required fields, templates, access groups).

  5. Ignoring third-party workflows.
    Fix: connect classification to third-party onboarding, contract clauses, and offboarding access revocation.

Enforcement context and risk implications

No public enforcement cases were provided in the source catalog for this requirement, but the risk is operationally real: without classification tied to OEM expectations, you can breach contractual confidentiality obligations, mishandle customer-controlled technical data, or fail a TISAX assessment because assessors cannot verify consistent implementation. (VDA ISA Catalog v6.0)

Practical 30/60/90-day execution plan

Use phases rather than calendar promises; adjust based on system complexity and number of OEMs.

First 30 days (foundation and alignment)

  • Identify domains, repositories, and top OEM partner requirements to map. (VDA ISA Catalog v6.0)
  • Draft four-level scheme with crisp definitions and examples.
  • Draft handling rules table and exception process.
  • Socialize with Engineering, IT, Legal/Commercial, and Procurement for fast feedback.

Days 31–60 (implementation in the places that matter)

  • Finalize OEM mapping matrix and get sign-off. (VDA ISA Catalog v6.0)
  • Implement minimum viable controls in core systems (templates, access groups, required metadata fields).
  • Launch targeted training for high-risk teams (engineering and procurement first).
  • Start spot checks and open remediation tickets for gaps.

Days 61–90 (evidence, repeatability, and audit readiness)

  • Expand implementation to remaining repositories and workflows.
  • Run a formal internal control check: sample artifacts across domains and verify classification + handling.
  • Tighten exceptions: expirations, compensating controls, approvals.
  • Package the evidence set for assessors: policy, mappings, configs, samples, training, checks. (VDA ISA Catalog v6.0)

Where Daydream fits (without adding process drag)

If you manage many third parties and OEM data-sharing paths, the work breaks down in tracking: who received what, under which classification, under what contractual constraints, and whether access was revoked. Daydream can serve as the system of record for third-party due diligence and ongoing oversight, linking data sharing requests, classification requirements, and evidence (approvals, attestations, access reviews) so audits stop being scavenger hunts.

Frequently Asked Questions

Do we have to use exactly “Public, Internal, Confidential, Strictly Confidential”?

VDA ISA 2.1.1 requires a scheme consistent with OEM partner requirements and industry standards, and the provided expectation includes at least those four levels. (VDA ISA Catalog v6.0) You can add internal nuance, but assessors will expect a clear mapping and consistent use.

What’s the difference between “Confidential” and “Strictly Confidential” in practice?

The difference must be visible in handling rules: tighter access approval, stricter sharing channels, and clearer limits on external distribution for “Strictly Confidential.” If both levels are handled the same way, your scheme will look theoretical.

How do we handle documents that contain mixed sensitivity (e.g., public brochure plus OEM-specific specs)?

Classify to the highest sensitivity content in the document or dataset, then segregate content if you need a lower-class shareable version. Add a rule that prohibits “partial sharing” unless a redacted or separated version is created and reviewed.

Do we need automated labeling tools to pass?

The requirement is to define and implement the scheme; automation can help but is not the only path. What matters is consistent application with evidence across key systems and workflows. (VDA ISA Catalog v6.0)

How do we extend classification to third parties and contractors?

Put the classification rules into onboarding and collaboration processes: contract/security exhibit alignment, approved sharing channels, and access revocation on offboarding. Keep an exception log for any nonstandard sharing arrangement.

What evidence is most persuasive to an assessor?

A tight package: the scheme and handling rules, an OEM mapping matrix with approvals, system screenshots showing enforcement points, labeled real examples from engineering and procurement, and spot-check results with remediation. (VDA ISA Catalog v6.0)

Frequently Asked Questions

Do we have to use exactly “Public, Internal, Confidential, Strictly Confidential”?

VDA ISA 2.1.1 requires a scheme consistent with OEM partner requirements and industry standards, and the provided expectation includes at least those four levels. (VDA ISA Catalog v6.0) You can add internal nuance, but assessors will expect a clear mapping and consistent use.

What’s the difference between “Confidential” and “Strictly Confidential” in practice?

The difference must be visible in handling rules: tighter access approval, stricter sharing channels, and clearer limits on external distribution for “Strictly Confidential.” If both levels are handled the same way, your scheme will look theoretical.

How do we handle documents that contain mixed sensitivity (e.g., public brochure plus OEM-specific specs)?

Classify to the highest sensitivity content in the document or dataset, then segregate content if you need a lower-class shareable version. Add a rule that prohibits “partial sharing” unless a redacted or separated version is created and reviewed.

Do we need automated labeling tools to pass?

The requirement is to define and implement the scheme; automation can help but is not the only path. What matters is consistent application with evidence across key systems and workflows. (VDA ISA Catalog v6.0)

How do we extend classification to third parties and contractors?

Put the classification rules into onboarding and collaboration processes: contract/security exhibit alignment, approved sharing channels, and access revocation on offboarding. Keep an exception log for any nonstandard sharing arrangement.

What evidence is most persuasive to an assessor?

A tight package: the scheme and handling rules, an OEM mapping matrix with approvals, system screenshots showing enforcement points, labeled real examples from engineering and procurement, and spot-check results with remediation. (VDA ISA Catalog v6.0)

Authoritative Sources

Operationalize this requirement

Map requirement text to controls, owners, evidence, and review workflows inside Daydream.

See Daydream
TISAX: Information Classification Scheme | Daydream