1. What happened
License covers Times, Cooking and Athletic content for Amazon answers and foundation-model training. Multi-year agreement; consideration not specified in this release.
2. Buyer / seller
Amazon is the AI buyer or model-development side. The New York Times Company owns or contributes the proprietary corpus.
3. Economics
No arrangement-specific consideration established in the inspected announcement. Economics scope: No data-license value is estimated. Aggregate corporate revenue, financing valuations, settlement payments and catalogue capacity are not treated as this arrangement’s price.. Strategic Signal has not estimated undisclosed consideration.
4. Proprietary data
Licensed New York Times, NYT Cooking and The Athletic content for Amazon answer experiences and foundation-model training. Historical depth: The exact licensed or transferred historical window is not disclosed.. Refreshability: Ongoing refresh obligations and the exact delivered corpus are not disclosed..
5. Rights and restrictions
permitted for Amazon foundation-model training under the announced license Ownership: not disclosed. Training: permitted for Amazon foundation-model training under the announced license. Post-training: not disclosed.
6. Why the data matters for AI
The disclosed arrangement supports model development using the authorized external corpus. The licensed content supports retrieval and source-grounded output. No recipient-accessible repeated context/intervention/outcome structure is established by this announcement.
7. Asset Score — Pending/100
Only documented AI use and bounded rights usability are assessed. Evidence for remaining comparative corpus or transaction factors is insufficient; coverage remains below the locked 70% threshold and no total score is published. Score coverage: 14%.
| Factor | Weight | Rating / 5 | Points |
|---|---|---|---|
| Uniqueness / scarcity | 12 | Unknown | — |
| Historical depth | 8 | Unknown | — |
| Decision → outcome richness | 13 | Unknown | — |
| Decision Value Density | 13 | Unknown | — |
| Real-world grounding | 10 | Unknown | — |
| Domain economic value | 9 | Unknown | — |
| Refreshability | 8 | Unknown | — |
| Proprietary advantage | 9 | Unknown | — |
| Model-learning usefulness | 10 | 4 | 8.0 |
| Non-replicability | 4 | Unknown | — |
| Rights usability | 4 | 3 | 2.4 |
8. Transaction Signal Score — Pending/100
Only documented AI use and bounded rights usability are assessed. Evidence for remaining comparative corpus or transaction factors is insufficient; coverage remains below the locked 70% threshold and no total score is published. Score coverage: 26%.
| Factor | Weight | Rating / 5 | Points |
|---|---|---|---|
| Disclosed economics | 16 | Unknown | — |
| Clean price discovery | 16 | Unknown | — |
| Separability of data consideration | 15 | Unknown | — |
| Explicit training / model-improvement use | 14 | 4 | 11.2 |
| Clarity of rights purchased | 12 | 3 | 7.2 |
| Competing bids | 10 | Unknown | — |
| Strategic buyer quality | 6 | Unknown | — |
| Repeatability / precedent value | 6 | Unknown | — |
| Independent model-uplift evidence | 5 | Unknown | — |
9. M&A / valuation implications
Analysis: this is evidence of a governed external-asset arrangement, not a standalone valuation comparable without separable economics.
10. Evidence boundary
Strict qualification rests on the defined external corpus and separately material licensed, research or acquired rights, rather than ordinary internal AI deployment. Multi-year agreement; consideration not specified in this release. Rights not expressly established remain not_disclosed. This historical addition is not a new November announcement.