Topics · Domain workflow

AI training and dataset audit

Auditors can check which dataset commitments and transformation claims were attached to a training or audit trail.

Problem

What breaks today

A model passes audit. Months later, a flag turns up in its outputs. The team needs to prove which dataset version the model actually saw, and which transformation pipeline produced it.

Mechanism

How ZK-SNAP responds

Receipts can bind dataset roots, pipeline stages, evaluation evidence, and training context into records that survive after the run.

Verifiable outcome

What a verifier can check

Auditors can check which dataset commitments and transformation claims were attached to a training or audit trail.

Related profiles and labels

Data trailTransformation trailDiscovery

Go deeper

Try the workflow, then read the spec.

Use Cases tells the story with cards. Proof Lab runs create and verify locally. Protocol holds the normative reference.

Use Cases Proof Lab How it works