The Rights
Foundry
Sign in
For AI and data teams

Training data you can put in front of procurement.

Rights-cleared music data for training, fine-tuning and evaluation. Consent on record for every track. Downstream commercial rights built into the license.

Talk to our AI team →Read the data standard
Rights manifest, excerpt (designed)
{ "snapshot": "cat_2027_01",
  "works": 35000,
  "consent": "genai_addendum_signed",
  "permitted": ["training", "fine_tuning",
                "evaluation", "generative_output"],
  "downstream_commercial": true,
  "territory": "worldwide" }
The procurement review

What your reviewers will ask, and our answers.

Every AI team now gets the same question from legal, from procurement and increasingly from regulators: what did you train on, and did you have the right to? The Rights Foundry is built so the answer is short, specific and documented: every licensed dataset is designed to ship with a rights manifest and a provenance report, so compliance becomes an output of the workflow, not a project.

✓Chain of titleRecording and composition cleared together for every track, from catalogs that hold signed artist agreements.
✓Artist consent for AI35,000+ tracks cleared for AI under signed GenAI Addenda. Opt-in, artist by artist.
✓Scope of usePer-track rights flags for training, fine-tuning, evaluation, generative output and style transfer.
✓Output rightsThe generative AI license includes downstream commercial rights explicitly.
✓TerritoryGlobally cleared.
✓Provenance and auditEvery licensed use is recorded. Versioned catalog snapshots, rights manifests and provenance reports are designed to ship at launch.
✓Regulatory supportRights records are structured to help model providers meet training data transparency and copyright policy expectations, including under Article 53 of the EU AI Act. Not legal advice.
✓Artist shareRights holders share in every AI deal on an ongoing basis. No flat-fee buyouts.
✓Data protection and securityData processing agreement available.
✓NeutralityIndependent and catalog-agnostic. Not owned by a label or a model company.
Data products

The AI Suite.

Training Corpus

AI-cleared, commercially released music with machine-readable metadata and rights flags.

Fine-Tuning Packs

Curated subsets by genre, mood, tempo and style.

Evaluation Datasets

Fixed, rights-cleared sets for reproducible benchmarking. Versioned snapshots at launch.

Retrieval and Embedding Layer

Similarity search and pre-computed embeddings at launch.

Licensed Output

Generative AI license with downstream commercial rights.

Try the format first

A gated set of 500 commercially released, ISRC-verified tracks with full metadata and rights flags is planned for Hugging Face in Q1 2027, under our Research and Evaluation License, alongside our open rights manifest schema.

License Collection One →

Priced by use-case risk

Pay for the rights your use actually needs.

Low

Search, classification, tagging, recommendation

Standard API access, usage-based pricing

Medium

Retrieval, music-aware features, non-generative integration

Expanded tier, annual agreement

High

Training, fine-tuning, generative output, style transfer

AI Suite agreement with generative AI license and downstream rights

How an engagement works

From scoping call to delivery, with every use on record.

Scoping callUse cases, volumes, delivery format and review requirements.
Sample and evaluation setWith metadata and rights flags, for your team to test.
AgreementStandard core rights terms; enterprise terms such as security, data processing and indemnity agreed to fit.
DeliveryEvery licensed use on record.
Why this is hard to copy

The asset is consent.

Anyone well funded can build an embedding index or scrape a dataset. Neither is the asset. The asset is consent: 35,000+ tracks under signed AI addenda, built on relationships with 3,500+ artists. That takes years and trust, not just capital.

Talk to our AI team.

Bring your use cases and your review checklist. We will bring the answers and a sample set.

Talk to our AI team →AI and data FAQ