Compliance Report

AI Chain-of-Custody Compliance Report

· Reporting Period:
Generated: · Schema Version:

1. Methodology

This report summarizes AI-assisted work attestations captured by AIAuth over the reporting period. All data is derived from cryptographically-signed receipts created by individual employees using the AIAuth Chrome extension or desktop agent.

What AIAuth captured: metadata only (timestamps, AI model name, file type, review decisions, chain linkage).

What AIAuth did not capture: content, prompt text, screen images, keystrokes, or any information that could identify the substance of the AI-assisted work.

Infrastructure: self-hosted AIAuth Enterprise server, operated by the reporting organization's IT team. Data resides on organization-controlled infrastructure throughout.

For auditors: every receipt ID in this report can be independently verified against the organization's public key using any AIAuth-compatible verifier, including the open-source reference implementation.

2. Attestation Volume

Total Attestations
Unique Attesters
AI Tools Used
File Attestations

3. Policy Violations — Severity Breakdown

SeverityCountResolvedOutstanding
Policies: dual-review-financial, no-rubber-stamping, external-must-attest, unverified-financial, shadow-ai-detected, ungoverned-ai-content. Definitions in Enterprise Admin Guide Section 4.7.

4. Department Compliance Scores

DepartmentAttestationsReview RateRubber-Stamp RateGrade

Grade: A (≥95% reviewed, 0 critical violations), B (≥85%), C (≥70%), D (≥50%), F (<50%).

5. Chain Integrity Summary

Total Document Chains
Complete Chains
Broken Chains
Integrity Rate

A "broken chain" means a receipt references a parent hash that cannot be found in the registry. This typically indicates a document whose lineage predates AIAuth adoption or whose parent receipt was created on a different AIAuth deployment.

6. Cross-Format Chain Verification

AIAuth's canonical-text hashing enables chain-of-custody preservation across format conversions (e.g., Excel → CSV → PDF). Receipts whose byte-level content hashes differ but share a canonical-text hash represent the same logical document in different formats.

Canonical Groups
Multi-Format Documents
AI-Authored Artifacts
Ungoverned AI Markers

7. Recommendations

  1. Loading…

8. Attestation

This report is generated by the AIAuth server. The metadata aggregations above are reproducible from the underlying enterprise_attestations table; the queries are documented in the AIAuth Enterprise Admin Guide, Appendix B.

Reviewed by (IT/Compliance Officer)
Date
AIAuth never sees, stores, or transmits your content. This report is derived from metadata only, produced by your organization's self-hosted AIAuth Enterprise server. Finch Business Services LLC (the software vendor) has no access to the data summarized here.