White paper WP-018
Cryptographic AI-BOM: ZK Provenance for Model Supply Chains
Prove where a model came from without handing over your weights.
At a glance
- Paper
- WP-018
- Topic
- Cryptographic AI-BOM: ZK Provenance for Model Supply Chains
- Format
- PDF + web summary
- Signatures
- ML-DSA-65 post-quantum (NIST FIPS 204)
- Sandbox
- Reproducible at affix-io.com/sandbox
- Company
- AffixIO, Wales, UK
Cryptographic AI-BOM: ZK Provenance for Model Supply Chains is an AffixIO technical paper. Prove where a model came from without handing over your weights.
Regulators want an AI bill of materials. Vendors cannot publish training sets or architecture diagrams. Zero-knowledge attestations let you prove dataset hashes, fine-tune steps, and evaluation gates were satisfied while keeping trade secrets private.
Summary
Regulators want an AI bill of materials. Vendors cannot publish training sets or architecture diagrams. Zero-knowledge attestations let you prove dataset hashes, fine-tune steps, and evaluation gates were satisfied while keeping trade secrets private.
Download the full PDF for technical detail, diagrams, and reproduction steps. Public sandbox: affix-io.com/sandbox.
Related reading
Frequently asked questions
What is an AI-BOM?
An AI bill of materials lists datasets, base models, fine-tuning runs, evaluation results, and deployment artefacts that produced a given model version.
Can ZK proofs hide proprietary training data?
Yes. You prove commitments to dataset hashes and licensing checks without revealing raw records or model weights.
Does this align with EU AI Act Article 11?
Cryptographic attestations support technical documentation requirements by providing verifiable supply chain records alongside human-readable docs.