| Quick facts | |
|---|---|
| Largest disclosed contract value | Reddit: $203.0M aggregate, 2–3 year terms[1] |
| Largest reported deal | News Corp–OpenAI: >$250M over five years (reported)[2] |
| Largest buyer budget | micro1: $1B over 12 months[3] |
| Cautionary data point | Shutterstock data offering −91% YoY, Q3 2026[4] |
Dataset
The tracker
| Date | Licensor → licensee | What was licensed | Value | Basis | Source |
|---|---|---|---|---|---|
| 2026-11 | Operating companies (program) → Polyshares | Slack/Teams, email, CRM, tickets, design files, call recordings, ERP exports | Typically $100K–$2M per licenseBuyer-published band; net 30–60 after approval and anonymization. | Buyer’s own published figure | [23] |
| 2026-10 | Operating companies (program) → micro1 | Enterprise operating data for RL environments | $1B over 12 months; published bands from $100K–$500KFinanced with capital provided by Citi and Hercules Capital. | Buyer’s own published figure | [3, 21] |
| 2026-10 | Trades, farming and manufacturing businesses (program) → HUMXN Research Labs | Service records, repair histories, job-site photos, wearable-camera footage | $50M initiative; ~$20 per approved hour of footageHomepros: “It remains unclear where the $50 million will come from.” | Buyer’s own published figure | [22, 20] |
| 2026-09 | Reddit → Licensing partners (not named) | Content licensing, reported inside “Other revenue” | $47.96M “Other revenue” in Q3 2026Reddit says substantially all licensing contract value comes from two partners whose terms are nearing their end. | Disclosed by the company | [6] |
| 2026-09 | Shutterstock → AI developers (not named) | Image, video and metadata libraries | $15.8M segment revenue in Q3 2026; data offering −91% YoYThe cautionary data point: commodity libraries are repricing. | Disclosed by the company | [4] |
| 2026-07 | Vitality Mechanical → micro1 | Labeling tasks documenting vendor invoices, credit applications, and payroll | ~$1,000 per task; >$150,000 expected in a quarterPer-task payments, not a purchase price for operating records. | Reported by press, unconfirmed | [19, 20] |
| 2026-03 | News Corp → Meta | U.S. and U.K. content for training and retrieval | Up to $50M a year, at least three yearsReported by the Wall Street Journal from people familiar with the matter. | Reported by press, unconfirmed | [8] |
| 2026-01 | Data partners (healthcare, media, audio, spatial) → Protege | Licensed datasets aggregated across hundreds of partners | $30M Series A extension; $65M total raisedCompany funding, not a payout figure; partners receive revenue share per use. | Disclosed by the company | [24] |
| 2025-05 | The New York Times → Amazon | Editorial content, including recipes and sports | $20M–$25M a yearAnnounced by The Times without terms; value reported by the Wall Street Journal. | Reported by press, unconfirmed | [10, 11] |
| 2024-11 | HarperCollins → Microsoft | Select nonfiction backlist titles, three-year term | $5,000 per title, split 50-50 with the authorOpt-in for authors; terms as described by the Authors Guild. | Disclosed by the company | [18] |
| 2024-05 | News Corp → OpenAI | Current and archived news content (WSJ, Barron’s, The Times, others) | More than $250M over five years, cash and creditsThe companies did not state a value; the figure is the Wall Street Journal’s reporting. | Reported by press, unconfirmed | [7, 2] |
| 2024-05 | Dotdash Meredith (IAC) → OpenAI | Magazine and web publishing content | At least ~$16M a year (fixed component)Derived from IAC’s Q3 2024 earnings call; a variable component is excluded. | Disclosed by the company | [9] |
| 2024-05 | Stack Overflow → OpenAI | Technical Q&A via OverflowAPI | Not disclosedAPI partnership. | Value not disclosed | [13] |
| 2024-05 | Taylor & Francis (Informa) → Microsoft | Non-exclusive access to advanced learning content | $10M+ initial access fee, recurring payments for three yearsInforma sold $75M of non-recurring data access in 2024 across Microsoft and one undisclosed partner (Bloomberg). | Disclosed by the company | [16, 17] |
| 2024-04 | Financial Times → OpenAI | FT journalism for training and attributed ChatGPT answers | Not disclosedStrategic partnership and licensing agreement. | Value not disclosed | [12] |
| 2024-04 | John Wiley & Sons → Large technology companies (not named) | Books and journal content | $23M AI licensing revenue in fiscal 2024; $40M in fiscal 2025Wiley’s fiscal year ends 30 April. Publishers Lunch reported $49M for fiscal 2026 from the company’s investor call. | Disclosed by the company | [14, 15] |
| 2024-02 | Reddit → Google | User posts and comments, via API access | About $60M a yearReported by Reuters from people familiar with the matter. | Reported by press, unconfirmed | [5] |
| 2024-01 | Reddit → Not identified in the quoted disclosure | User posts and comments | $203.0M aggregate contract value, 2–3 year termsArrangements entered in January 2024, disclosed in the IPO prospectus. | Disclosed by the company | [1] |
Cases a data seller’s counsel will cite
| Case | Court / filed | Status | Why it matters to a seller | Source |
|---|---|---|---|---|
| The New York Times Co. v. Microsoft Corp. & OpenAI | S.D.N.Y., 2023-12-27 | Pending | Unlicensed training on premium content is being litigated by its owners — the reason licensed, documented data commands a price. | [25] |
| Thomson Reuters v. Ross Intelligence | D. Del., 2020-05 | Fair use rejected on summary judgment (non-generative AI) | The first U.S. ruling that training on someone else’s proprietary records was not fair use. Proprietary operating records are the closest analogue. | [26] |
| Bartz v. Anthropic PBC | N.D. Cal., 2024-08 | $1.5B class settlement approved; about $3,000 per book | How material was acquired mattered as much as how it was used. Chain of title is a priced risk. | [27] |
| Kadrey v. Meta Platforms | N.D. Cal., 2023-07 | Summary judgment for Meta on the fair use question as argued | Outcomes turn on the record each plaintiff builds, which is why buyers prefer licensed data to litigation risk. | [28] |
| Reddit, Inc. v. Anthropic PBC | San Francisco Superior Court, 2025-06-04 | Pending | Contract terms, not only copyright, govern who may use data — the same logic applies to your vendors’ terms over your records. | [29] |
Section 2
How to read the tracker
The basis column is the point. A figure in an SEC filing or an earnings call is not the same kind of fact as a number a newspaper got from “people familiar with the matter,” and neither is the same as a band a buyer prints on its own marketing page. Most trackers flatten all three into one column. This one does not.
| Basis | Meaning | Example |
|---|---|---|
| Disclosed by the company | Stated by the licensor in a filing, release or earnings call | Reddit’s $203.0M aggregate contract value[1] |
| Reported by press, unconfirmed | Press reporting the parties have not confirmed | News Corp–Meta, up to $50M a year[8] |
| Buyer’s own published figure | A buyer’s advertised band or budget | micro1’s bands and $1B budget[21] |
| Value not disclosed | The deal is public; the money is not | Financial Times–OpenAI[12] |
Section 3
What the record shows
Publishers were first, and their deals are the only ones with sizeable public numbers. News, scholarly and platform content dominate the disclosed and reported figures — Wiley alone reported $23 million of AI licensing revenue in fiscal 2024 and $40 million in fiscal 2025.[14]
Concentration and decline are on the record too. Reddit says substantially all of its licensing contract value comes from two partners nearing the end of their terms[6]; Shutterstock’s data offering fell 91% year over year in Q3 2026.[4]
Operating-company data is the 2026 story, and it has almost no disclosed prices. The money is in buyer budgets and advertised bands — micro1’s $1 billion commitment[3], Polyshares’ $100K–$2M range[23] — and one reported per-task example in HVAC.[20]
Section 4
Method and corrections
An entry is included only when we could open the source and read the figure in it. Values are quoted in the source’s own terms; we do not annualise, convert or combine. Reviewed 16 November 2026. To report a deal or a correction, use the contact form with a link to the source.
Frequently asked questions
What are the biggest AI data licensing deals?
How much does Reddit make from AI data licensing?
How much has Shutterstock made from AI data?
Can I download this dataset?
References
Sourced from primary industry reports, government data, SEC filings, and peer-reviewed valuation research. Retrieved at time of publication.
- 1.Reddit, Inc. — Form S-1 Registration Statement (February 2024) — “In January 2024, we entered into certain data licensing arrangements with an aggregate contract value of $203.0 million and terms ranging from two to three years.” (2024)
- 2.Associated Press — OpenAI to start using news content from News Corp. as part of a multiyear deal — Notes the companies would not discuss value; the Wall Street Journal reported it could be worth more than $250 million over five years (2024)
- 3.micro1 — micro1 announces commitment to spend $1B on enterprise data acquisition over the next 12 months — Announcement of 9 October 2026, financed with capital from Citi and Hercules Capital; also issued via Business Wire (2026)
- 4.Shutterstock, Inc. — Form 10-Q for the quarterly period ended September 30, 2026 — Data, Distribution and Services segment revenue and the decline in the metadata data offering (2026)
- 5.Reuters — Exclusive: Reddit in AI content licensing deal with Google — Reported the Google agreement at about $60 million a year, citing people familiar with the matter; 22 February 2024 (2024)
- 6.Reddit, Inc. — Form 10-Q for the quarterly period ended September 30, 2026 — Content licensing revenue, partner concentration, and the scraping risk factor (2026)
- 7.News Corp — News Corp and OpenAI Sign Landmark Multi-Year Global Partnership — Company announcement of 22 May 2024; financial terms not stated (2024)
- 8.The Wall Street Journal — News Corp, Meta in AI Content Licensing Deal Worth Up to $50 Million a Year — Multiyear agreement of at least three years covering U.S. and U.K. content, citing people familiar with the matter; March 2026 (2026)
- 9.Adweek — OpenAI Is Paying Dotdash Meredith At Least $16 Million to License Its Content — Derived from IAC’s third-quarter 2024 earnings call; the figure is the fixed component only, November 2024 (2024)
- 10.The New York Times — The Times and Amazon Announce an A.I. Licensing Deal — Company announcement of 29 May 2025; financial terms not stated (2025)
- 11.The Wall Street Journal — Amazon to Pay New York Times at Least $20 Million a Year in AI Deal — Reported $20 million to $25 million a year, citing people familiar with the matter (2025)
- 12.Financial Times — Financial Times announces strategic partnership with OpenAI — Strategic partnership and licensing agreement announced April 2024; financial terms not disclosed (2024)
- 13.Stack Overflow — Stack Overflow and OpenAI Partner to Strengthen the World’s Most Popular Large Language Models — API partnership via OverflowAPI announced 6 May 2024; financial terms not disclosed (2024)
- 14.John Wiley & Sons — Research Growth, AI Licensing, and Cost Reduction Drive Wiley’s Fiscal 2025 Results — “$40 million in total AI licensing revenue realized in Fiscal 2025 compared to $23 million in Fiscal 2024” (2025)
- 15.Publishers Lunch — Wiley Expects AI Sales to Multiply — Reports $49 million of AI licensing revenue in Wiley’s fiscal 2026, from the company’s investor call; June 2026 (2026)
- 16.Inside Higher Ed — Taylor & Francis AI deal sets “worrying precedent” — Non-exclusive access for Microsoft; Informa reported an initial $10 million access fee followed by recurring payments over three years (2024)
- 17.Bloomberg — Academic Publishers Sign AI Deals as Trump Cuts Research Funding — Informa sold $75 million of non-recurring data access in 2024 through its Microsoft partnership and another undisclosed partner; 10 June 2025 (2025)
- 18.The Authors Guild — HarperCollins AI Licensing Deal — A $5,000 fee per title, split 50-50 between a participating author and HarperCollins, for a three-year AI training license; opt-in for authors (2024)
- 19.The Information — Why Data Labeling Startups Are Paying HVAC Companies $150,000 For Data — Original report of the HVAC example, by Stephanie Palazzolo, 29 July 2026 (subscription required) (2026)
- 20.Homepros — Tech firms’ hunt for data reaches HVAC, plumbing contractors — Trade-press report of 17 October 2026 restating the HVAC example and quoting a buyer-side founder on how long the market may last (2026)
- 21.micro1 — Monetize your company’s operational data — Company Data Partnerships program page — A buyer’s own published payout bands, eligibility criteria, payment terms and data-handling commitments; marketing claims, not audited outcomes (read November 2026)
- 22.HUMXN Research Labs — HUMXN Research Labs Wants to Buy $50 Million Worth of Skilled Trades and Manufacturing Data for AI & Robotics Training — Company release of 13 October 2026, also distributed by Newsfile (2026)
- 23.Polyshares — Sell your company’s operational data — program page — A buyer’s own published deal size, terms, payment timing and anonymization statements; marketing claims, read 16 November 2026 (read November 2026)
- 24.Protege — Protege Raises $30 Million Led by a16z to Unlock Access to Data for AI Development — Announcement of 8 January 2026; total funding of $65 million and revenue-share payouts to data partners (2026)
- 25.The New York Times — New York Times Sues OpenAI and Microsoft Over Use of Copyrighted Work — Complaint filed 27 December 2023 in federal court in Manhattan (2023)
- 26.Reuters — Thomson Reuters wins AI copyright “fair use” ruling against one-time competitor — Judge Stephanos Bibas, D. Del., 11 February 2025 — the first U.S. ruling on fair use in AI training (2025)
- 27.Associated Press — Judge approves a $1.5B Anthropic settlement over books used to train its chatbot — Bartz v. Anthropic, N.D. Cal.; about $3,000 per book for works obtained from pirate libraries (2025)
- 28.U.S. Copyright Office — Fair use index summary — Kadrey v. Meta Platforms, Inc., 788 F. Supp. 3d 1026 (N.D. Cal. 2025) — Summary judgment for Meta on the fair use question presented, 25 June 2025 (2025)
- 29.Associated Press — Reddit sues Anthropic, accusing the AI company of illegally “scraping” user comments — Filed 4 June 2025 in San Francisco Superior Court; contract and unfair-competition claims rather than copyright (2025)