FIUND AI Data Licensing Index

Methodology

Unit of analysis: one publicly reported agreement between a content or data licensor and an AI company. A renewed or materially expanded agreement may be a separate row if the public reporting treats it as a new deal.

Inclusion rule: the agreement must permit AI model training, retrieval/search use, or a licensed AI product. Announced negotiations without a completed agreement are excluded unless the row explicitly describes the reporting status.

Confidentiality rule: no FIUND customer, supplier, contributor, partner, prospect, contract, or private commercial relationship is sourced from internal records or added to this index. Companies appear only as parties to independently reported public-market agreements.

Terms: figures are recorded only when a cited public source reported them. “Undisclosed” means comparable commercial terms were not public; estimates are kept in the source’s language and are not converted into a FIUND price benchmark.

Taxonomy: content type and use scope are normalized for comparison. The CSV preserves the source-linked summary so readers can inspect context that a category cannot capture.

Maintenance: new rows require a public source, a stable deal ID, and editorial review. Corrections update the version date; prior files can be requested for audit.

What this index can and cannot show

The index shows that licensing is happening across multiple media categories and that many commercial terms remain private. It does not estimate the full market, infer undisclosed prices, or treat reported headline values as comparable unit economics.

Current records

The downloadable file contains 32 rows. For a human-readable view with deal summaries and direct citations, use the AI content-licensing deals tracker.