Deals tracker

AI content-licensing deals tracker

A running, cited record of the public deals where content and data owners licensed their material to AI companies — news publishers, stock libraries, forums, book and academic publishers, and music rights holders. It exists to make one thing legible: content is being paid for, and the market has prices.

Lawsuit tracker Licensing guides Weekly news

Every row below is a real, publicly reported agreement, sorted newest first. Where a source reported a deal value it is shown and marked as reported — these are public market facts, not fiund’s pricing. Where terms were never disclosed, the entry reads “undisclosed.” We track 31 deals across 5 content types: News & journalism, Music, Books & academic, Forum/UGC, Stock imagery.

Publicly reported deals only. Terms are as reported by cited sources; “undisclosed” where not made public. Reported figures are attributed to their source and may cover a single year or a multiyear total, as noted.

AnnouncedLicensorAI companyContentScopeTermsSource
Mar 2026News CorpMetaNews & journalism
News Corp agreed to license copyrighted content from its US and UK titles to Meta for AI use, in a multiyear deal reported to run at least three years.
Model trainingUndisclosed; reported at up to $50M per year, multiyearEditor & Publisher · Yahoo Finance
Nov 2025Warner Music GroupSunoMusic
Warner Music Group settled its copyright litigation with Suno and licensed its catalog for a new generation of licensed AI music models, with Suno also acquiring the Songkick concert platform from WMG.
Model training + productUndisclosed (settlement plus licensing; Suno acquired Songkick from WMG)PR Newswire · Billboard
Oct 2025Universal Music GroupUdioMusic
Universal Music Group and Udio settled UMG’s copyright suit and struck licensing agreements to power a new, licensed AI music platform trained on authorized recordings and publishing from participating artists.
Model training + productUndisclosed (settlement plus licensing)PR Newswire · Billboard
May 2025The New York TimesAmazonNews & journalism
The New York Times signed its first AI licensing deal, letting Amazon use Times journalism, NYT Cooking recipes, and The Athletic across Amazon’s AI products, including Alexa.
Training + search/RAGUndisclosed; reported at about $20M–$25M per yearAxios · GeekWire
Feb 2025The GuardianOpenAINews & journalism
The Guardian licensed its journalism to OpenAI for use in ChatGPT with attribution and links, and adopted ChatGPT Enterprise internally.
Training + search/RAGUndisclosedPress Gazette · Nieman Lab
Nov 2024HarperCollinsMicrosoftBooks & academic
HarperCollins became the first Big Five publisher to license select nonfiction titles for AI training, offering participating authors an opt-in split of a reported $5,000-per-book fee.
Model trainingReported at $5,000 per title over three years, split with authorsAuthors Guild · Transparency Coalition
Nov 2024AxiosOpenAINews & journalism
OpenAI partnered with Axios in a three-year deal that funds four new local newsrooms and gives OpenAI access to Axios content to enhance GPT search answers with sourced links.
Search/attributionUndisclosed; three-year deal partly funding four new Axios newsroomsOpenAI · eMarketer
Oct 2024HearstOpenAINews & journalism
Hearst licensed content from more than 20 magazine brands and 40 newspapers, including Cosmopolitan, Esquire, and the San Francisco Chronicle, to OpenAI with citations and source links.
Training + search/RAGUndisclosedOpenAI · CNBC
Aug 2024Condé NastOpenAINews & journalism
Condé Nast agreed to let OpenAI train on and surface content from titles including Vogue, The New Yorker, Wired, and Vanity Fair across ChatGPT and the SearchGPT prototype.
Training + search/RAGUndisclosedOpenAI · TechCrunch
Aug 2024WileyUndisclosed large tech companyBooks & academic
Academic publisher Wiley licensed access to scholarly content for AI training, disclosing that it expected roughly $44 million from AI partnerships, without an author opt-out.
Model trainingUndisclosed; company disclosed about $44M expected from AI dealsbeSpacific · Books+Publishing
Jul 2024Taylor & FrancisMicrosoftBooks & academic
Academic publisher Taylor & Francis licensed research content to Microsoft for AI training, a deal that drew criticism from authors who said they were not consulted or paid extra.
Model trainingUndisclosed; reported at about $10M (initial)Inside Higher Ed · The Bookseller
Jun 2024TimeOpenAINews & journalism
Time licensed its 101-year editorial archive and current content to OpenAI for use in its products, with citations and links back to Time.com.
Training + search/RAGUndisclosedAxios · The Hollywood Reporter
May 2024News CorpOpenAINews & journalism
News Corp licensed content from titles including The Wall Street Journal and the New York Post to OpenAI, which can display the content in ChatGPT responses and use it in its products.
Training + search/RAGReported at over $250M across five years (cash plus OpenAI credits)Variety · Nieman Lab
May 2024Vox MediaOpenAINews & journalism
Vox Media licensed content from titles including The Verge, Vox, and New York Magazine to OpenAI, gaining brand attribution and audience referrals in ChatGPT.
Training + search/RAGUndisclosedAxios · Variety
May 2024The AtlanticOpenAINews & journalism
The Atlantic licensed its content to OpenAI for use in ChatGPT and gained access to OpenAI technology for an experimental product effort called Atlantic Labs.
Training + search/RAGUndisclosedAxios · Variety
May 2024Dotdash MeredithOpenAINews & journalism
Dotdash Meredith licensed content from brands including People, Better Homes & Gardens, and Investopedia to OpenAI, and adopted OpenAI models for its contextual ad product.
Training + search/RAGUndisclosed; reported at a minimum of about $16M per yearAdweek · PR Newswire
May 2024RedditOpenAIForum/UGC
Reddit licensed real-time access to its Data API to OpenAI for training and for surfacing Reddit content in ChatGPT, and adopted OpenAI-powered features on Reddit.
Training + search/RAGUndisclosed; reported at about $70MOpenAI · CNBC
May 2024Stack OverflowOpenAIForum/UGC
Stack Overflow licensed its vetted technical Q&A to OpenAI through its OverflowAPI, to improve model performance on coding tasks.
Model trainingUndisclosedVentureBeat · TechRadar
Apr 2024Financial TimesOpenAINews & journalism
The Financial Times licensed its journalism to OpenAI for training and for attributed use in ChatGPT, and became an OpenAI product customer.
Training + search/RAGUndisclosed; reported at an estimated $5M–$10MDigiday · Quartz
Apr 2024ShutterstockAppleStock imagery
Shutterstock licensed millions of images to Apple for AI training, one of several Shutterstock deals with big-tech buyers including Meta, Google, and Amazon.
Model trainingUndisclosed; reported at roughly $25M–$50MVentureBeat · 9to5Mac
Apr 2024FreepikUndisclosed AI companiesStock imagery
Stock-image marketplace Freepik signed agreements to license the bulk of its archive of roughly 200 million images to two large tech firms for AI training, at a reported rate near three cents per image.
Model trainingUndisclosed partners; reported at about $0.03 per imagePetaPixel
Apr 2024PhotobucketMultiple AI companies (reported)Stock imagery
Photobucket updated its terms to allow AI licensing and was reported to be negotiating with multiple tech companies to license its roughly 13 billion photos and videos for training image models.
Model trainingUndisclosed; reported in negotiations, with per-image rates reportedPetaPixel
Mar 2024Le MondeOpenAINews & journalism
French newspaper Le Monde licensed its content to OpenAI to train models and surface French-language news in ChatGPT.
Training + search/RAGUndisclosedOpenAI · Bloomberg
Mar 2024Prisa MediaOpenAINews & journalism
Spanish media group Prisa Media licensed content from titles including El País to OpenAI to train models and bring Spanish-language news to ChatGPT.
Training + search/RAGUndisclosedOpenAI · Bloomberg
Feb 2024RedditGoogleForum/UGC
Reddit licensed its content to Google for AI model training and for use in Google products via expanded API access, in a deal reported at roughly $60 million per year.
Training + search/RAGReported at about $60M per yearAOL / Reuters · CNBC
Feb 2024Stack OverflowGoogleForum/UGC
Stack Overflow made Google its first OverflowAPI customer, licensing its public developer knowledge base to train and ground Google’s Gemini models for Google Cloud.
Model trainingUndisclosedTechCrunch · InfoWorld
Feb 2024Automattic (Tumblr & WordPress.com)OpenAI and MidjourneyForum/UGC
Automattic was reported to be preparing to share public Tumblr and WordPress.com content with OpenAI and Midjourney for AI training, with an opt-out for users.
Model trainingUndisclosed404 Media · Engadget
Dec 2023Axel SpringerOpenAINews & journalism
Axel Springer licensed content from titles including Politico, Business Insider, and Bild to OpenAI for model training and for summarized, attributed answers in ChatGPT.
Training + search/RAGUndisclosed; reported in the tens of millions of euros over about three yearsTechCrunch · The Decoder
Sep 2023Getty ImagesNVIDIA (technology partner)Stock imagery
Getty Images licensed its own fully owned library to build a commercially safe generative image model with NVIDIA, and set up a system to pay contributors whose work trained the model.
Model trainingUndisclosed; contributor royalty pool for training imageryNVIDIA · Getty Images
Jul 2023The Associated PressOpenAINews & journalism
The Associated Press licensed part of its news archive to OpenAI in one of the first publisher deals, giving OpenAI access to text stories in exchange for access to OpenAI technology and expertise.
Model trainingUndisclosedThe Washington Post · VentureBeat
Jul 2023ShutterstockOpenAIStock imagery
Shutterstock expanded its relationship with OpenAI into a six-year agreement to license images, video, music, and metadata for AI training, with a contributor compensation fund.
Model trainingUndisclosed; Shutterstock reported $104M in AI-licensing revenue across partners in 2023PetaPixel · Yahoo Finance

Own content and want to license it? See how to sell to AI companies. Building a model and need rights-cleared material? Browse licensed data.

Keep reading

Sitting on content AI companies want?

These deals are the headline names. The same demand runs down the long tail of audio, video, and archives — often with rights already in hand. The first conversation is short and specific.

Talk about licensing