Audit AI data rights provenance integrity

What it's for

Gives leaders a defensible AI data bill of materials: which systems used which data, under which grant and purpose, through which derivative chain—and which uses must be quarantined now.

What you give it

Inputs split into evidence read from your connected systems, calibration your team owns, and numerical controls that affect precision but never the result's meaning.

{
  "data_assets": [
    {
      "acquired_at_ms": 100,
      "content_hash": "sha256:support-knowledge-snapshot",
      "deletion_requested_at_ms": null,
      "evidence_verified": true,
      "id": "support-knowledge-snapshot",
      "parent_asset_ids": [],
      "rights_grant_id": "grant-support-eu",
      "subject_consent_required": false,
      "subject_consent_verified": false
    }
  ],
  "rights_grants": [
    {
      "derivative_use_allowed": true,
      "evidence_verified": true,
      "id": "grant-support-eu",
      "permitted_jurisdiction_ids": [
        "eu"
      ],
      "permitted_purpose_ids": [
        "customer-support"
      ],
      "permitted_use_types": [
        "retrieval"
      ],
      "revoked_at_ms": null,
      "rights_holder_id": "customer-contract-pool",
      "sublicensing_allowed": true,
      "valid_from_ms": 0,
      "valid_until_ms": 10000
    }
  ],
  "usage_events": [
    {
      "asset_id": "support-knowledge-snapshot",
      "derivative_created": false,
      "evidence_verified": true,
      "id": "retrieval-use-1",
      "jurisdiction_id": "eu",
      "occurred_at_ms": 1000,
      "output_persisted": true,

What you get back

This is the actual output of running the example above — computed by the same function the platform calls, not an illustration.

{
  "assumptions": [
    "Grant, consent, revocation, deletion, purpose, jurisdiction and lineage records are authoritative, immutable and joined by stable tenant-scoped IDs.",
    "Every persisted output is conservatively treated as a transferable derivative requiring the declared right; local counsel owns the actual legal interpretation.",
    "Passing proves consistency with represented grants, not ownership, consent validity, copyrightability, regulatory compliance or permission to deploy."
  ],
  "counts": {
    "data_assets": 1,
    "exposed_systems": 0,
    "invalid_usage_events": 0,
    "rights_grants": 1,
    "usage_events": 1,
    "valid_usage_events": 1
  },
  "decision": "rights_provenance_verified",
  "evidence": {
    "gate_pass": true,
    "minimum_required_fraction": 0.95,
    "verified_fraction": 1
  },
  "failure_counts": {},
  "integrity_pass": true,
  "limitations": [
    "The audit cannot discover unregistered shadow datasets, memorized content, fraudulent attestations or rights restrictions absent from the supplied ledger.",
    "Results concern data assets and systems, never the character, nationality, loyalty or employment suitability of a contributor or rights holder."
  ],
  "method": "ai_data_rights_lineage_purpose_integrity_audit_v1",
  "truncated_detail_rows": 0,
  "usage_diagnostics": []
}

How it works

Statistical audit & measurement — Check whether a number is fit to decide on: coverage, timing, reconciliation, and the gaps a dashboard hides.

1. 1 Validate unique grants, canonical content hashes, authoritative consent/deletion state and an acyclic parent-asset lineage.
2. 2 For every usage event, traverse the full ancestor set and test grant window, revocation, purpose, use type, jurisdiction, consent, deletion and derivative rights.
3. 3 Quarantine unverified uses, aggregate failure mechanisms and affected systems, and enforce a portfolio evidence-coverage gate.

Before you trust it

Every tool in the catalog ships with the conditions under which its answer is meaningful — and the conditions under which it should abstain instead of guessing.

Calibrating it to your org

Calibration workflow

1. 1 Define the management decision, target outcome, aggregate unit, privacy boundary, cadence, and prediction/intervention horizon for this organization.
2. 2 Build a tenant-scoped historical cohort using only information available before each prediction or decision; preserve zero periods, censoring, assignment probabilities, and unresolved outcomes when the method requires them.
3. 3 Estimate statistical parameters on training history, but obtain costs, utilities, risk tolerance, practical-effect thresholds, capacity, and policy constraints from accountable decision owners.
4. 4 Validate on later time windows or held-out aggregate units at the deployment grain, against a simple baseline and the function-specific validation strategy.
5. 5 Deploy only if the returned decision clears evidence, overlap, calibration, robustness, and guardrail checks; warning, unsupported, schema-gap, and fallback decisions are abstentions.
6. 6 Monitor realized outcomes, data drift, coverage, and decision regret; recalibrate at a governed cadence or after a detected regime/definition change, never merely because a stakeholder dislikes the result.

Call it from your AI

You don't wire up 388 tools in your MCP client. The GitRevio MCP server exposes 18 tools, three of which let an agent search the catalog, read a tool's schema, and run it — so the assistant finds this one on its own.

gitrevio_capabilities_search
  { "q": "audit every ai training finetuning retrieval" }
  → finds "audit_ai_data_rights_provenance_integrity"

gitrevio_capability_describe
  { "capability_id": "audit_ai_data_rights_provenance_integrity" }
  → returns the input schema and agent guidance shown on this page

gitrevio_capability_run
  { "capability_id": "audit_ai_data_rights_provenance_integrity", "arguments": { ... } }
  → returns the result shown above

Works in Claude Desktop, Claude Code, Cursor, Cline, Continue.dev, Goose and Aider. See the MCP server.

Related tools

See every tool in AI risk, rights & assurance →