Driving the news:AI industry groups are urging an appeals court in the United States to block what they call the largest copyright class action ever certified, warning it could “financially ruin” the industry.
The lawsuit, brought by three authors over Anthropic’s AI training, could swell to seven million claimants, each eligible for up to $150,000 in damages.
Anthropic says that’s hundreds of billions in potential liability, a “death knell” four months before trial.
Catch up fast:
U.S. District Judge William Alsup certified the massive class, relying, Anthropic claims, on “50 years” of experience instead of a “rigorous analysis” of membership and ownership questions.
Anthropic is petitioning the appeals court to reverse the decision, warning that the pressure to settle could wipe out its defenses and set precedent for other AI lawsuits.
Why it matters:
Industry groups say allowing copyright class actions in AI training cases will chill investment, slow innovation, and jeopardise U.S. AI ‘leadership’.
“The technology industry cannot withstand such devastating litigation,” wrote the Consumer Technology Association and Computer & Communications Industry Association.
Between the lines:
Authors’ groups, including the Electronic Frontier Foundation, American Library Association, and Authors Alliance, also back Anthropic’s appeal.
They argue copyright suits are a poor fit for class actions:
Determining rights for millions of works is messy, orphan works, partial rights, defunct publishers, and deceased authors complicate ownership.
The court’s opt-out plan risks excluding authors who never hear about the case.
Google’s past $34.5m effort to build a “Books Rights Registry” shows how costly identifying rights holders can be.
The bottom line: Both AI industry advocates and some author groups warn the current case risks forcing settlements without answering fundamental questions about AI training and copyright, leaving everyone in the dark.
The rise of artificial intelligence has sparked intense debates about copyright, intellectual property, and the role of human creativity in the digital age.
What’s at play? As AI firms train their models on vast amounts of data – some of which may include copyrighted materials – legal disputes have emerged.
However, rather than engaging in costly and prolonged court battles, copyright holders and AI companies should focus on collaboration.
Similarly, artists and authors have sued AI firms like Stability AI and Midjourney, arguing that their work was used without proper credit or compensation.
In another case, Getty Images sued Stability AI for allegedly using millions of copyrighted images to train its Stable Diffusion model without authorisation. The lawsuit claimed that AI-generated outputs could even contain remnants of Getty’s watermark, underscoring concerns about intellectual property rights and fair use.
These cases illustrate the growing conflict between AI development and content ownership.
However, litigation is often a slow and expensive process with uncertain outcomes.
A more constructive approach would be for AI firms and copyright holders to work together to establish mutually beneficial agreements.
The need for high-quality source data
AI models rely on vast amounts of data, and the quality of that data directly impacts their performance.
Poor-quality or unauthorised training data can lead to biased, inaccurate, or legally problematic outputs.
“As AI-generated content becomes more prevalent, there is an increasing need for reliable, high-quality source material,” says Stephen Conmy, founder of Contenseo.
This reality creates an opportunity for copyright holders and AI firms to negotiate agreements that benefit both parties.
Rather than scraping content without permission, AI companies could license high-quality materials from publishers, authors, and artists.
In return, content creators could receive fair compensation and recognition while ensuring their work contributes to AI development in a controlled and ethical manner.
How AI firms can support and develop high-quality content creators
If AI firms want their tools to remain relevant and valuable, they must invest in the creators who produce the original works that feed their systems. Here are several ways they can do so:
Revenue-sharing agreements: AI firms can establish licensing deals with publishers, authors, artists, and musicians to fairly compensate them for their work. For example, rather than relying on scraped text from books, AI firms could pay publishers to access their archives legally.
Creator partnerships: AI companies can work directly with artists, writers, and musicians to create AI-generated works in collaboration rather than competition. For instance, AI tools could assist authors in writing, musicians in composing, and visual artists in designing, with clear acknowledgment of their contributions.
Content verification and attribution : AI firms could implement tools to track and verify original content used in training datasets. This would help ensure that creators receive credit and compensation when their work is used to develop AI models.
Funding and grants for original content: Some AI firms have started funding journalism and creative projects. OpenAI, for example, has provided grants for researchers studying the ethical use of AI. Expanding these initiatives to include funding for artists, writers, and other creatives could foster goodwill and encourage innovation.
Better transparency in AI training data: One of the biggest concerns for copyright holders is that they don’t know if their work is being used without permission. AI companies should offer more transparency about their training datasets and allow content creators to opt in or out.
The future: Cooperation over confrontation
AI is here to stay, and its development will only accelerate in the coming years.
Instead of seeing AI as a threat, copyright holders can look at the potential benefits of collaboration.
AI-generated content can amplify human creativity rather than replace it, provided that content creators are fairly compensated and credited.
On the other hand, AI firms must acknowledge that their models are only as good as the data they are trained on.
Without human-created content, these models risk producing generic, low-quality, or even legally dubious results.
The first major copyright trial of the generative AI industry has begun in London.
Getty Images versus Stability AI is a courtroom battle that could decide the future relationship between copyright holders and generative AI firms.
Getty claims that Stability AI, while developing its image maker Stable Diffusion, took part in ‘brazen infringement’ of Getty’s image collection ‘on a staggering scale’.
‘What Stability did was inappropriate,’ said Getty CEO Craig Peters in 2023.
He said copyright holders, creators and owners of intellectual property should grant permission to any AI firm that wishes to use their works to train their models.
Lawyers for the AI image maker argue that Getty’s claims ‘represent an overt threat to Stability’s entire business, and the wider generative AI industry’ and that ‘only a tiny proportion’ of the outputs of its AI image-generator ‘look at all similar’ to images stored on Getty’s databases.
Stephen Conmy, editorial director of Contenseo, says the trial is very significant. ‘The AI developers want copyright exemptions to train their AI models, which remains unlikely to happen. However, the decision of the trial judge will strengthen the hand of either side in the ongoing, worldwide legal battles between copyright holders and AI developers.’
Interesting fact
Sean Parker, former president of Facebook, is chair of Stability’s board and a significant investor in the company. Parker is very familiar with copyright violation court cases, having been the co-founder of the music streaming site Napster, which was sued by Metallica over copyright violations.
ProRataAI, developer of Gist.ai, the first AI search engine built entirely on licensed content, now has over 500 licensed publications powering Gist.ai’s answers.
Recent content licensing deals include Fast Company, The New Republic,The Philadelphia Inquirer, Boston Globe Media, New York Magazine, Frommer’s, Reader’s Digest, The Nation, Newsday, and Tom’s Guide.
“Gist.ai answers every query using 100 percent licensed content so consumers get authoritative, accurate answers and publishers share in the value their important journalism creates,” says Bill Gross, CEO of ProRata (pictured).
“We prove that generative AI can both honor creators and deliver an outstanding user experience.”
Compensation for copyright holders
ProRata’s patent-pending attribution system analyses AI-generated content, identifies the value provided by individual sources, and calculates proportional compensation for contributors.
“AI driven search does not have to be based on theft; publishers can, and should, be compensated for the use of their copyrighted material,” said Pauline Frommer, co-president of Frommer Media LLC. “That’s the ethical business model ProRata is pioneering, and why we at Frommer’s, are proud to partner with them.”
Andrew Min, VP, Global Partnerships and Licensing at publishing house Future, says, “Our collaboration with Gist.ai allows us to distribute our specialist content across new and innovative platforms, helping more people find accurate information from sources they can trust.”
High-quality content deals
Other content licensing deals with Gist.ai include Vox Media (owner of New York Magazine, The Verge, Vox, and Eater), Spanish news agency EFE, Spanish-language publisher PRISA Media, and various publications from Johns Hopkins University Press.
ProRata also recently introduced ProRata Ads, an advertising platform leveraging high-quality content to embed relevant ads within AI-generated answers. Early tests on partner websites have shown increases in click-through rates and user engagement, says ProRata.
About ProRata
Founded in 2024 by Bill Gross, ProRata develops ethical AI solutions that ensure content owners are credited and compensated for the use of their work. The firm’s leadership team includes former executives and engineers from Microsoft, Google, and Meta. ProRata is backed by investors and partners, including, Mayfield Fund, Revolution Ventures, Prime Movers Lab, dmg media, BOLD Capital, XPV-Exponential Ventures, Calibrate Ventures, MVP Ventures, and Idealab Studio. More information is available at ProRata.ai.
Why it matters: Large language model (LLM) creators are now willing to pay for reliable data feeds rather than risk using copyrighted text without consent.
AI developers who want access to high-quality, up-to-date data are finding it prudent to negotiate licenses, rather than face litigation and brand damage.
Many high profile licensing deals show that paying for data is becoming a norm.
We are witnessing the foundation of a new content licensing industry for AI, where deals between AI firms and content owners become commonplace.
Given the continuous creation of new content, licensing by AI firms will be an ongoing frontier.
The key point is that much of the best human knowledge (from literature to scientific research) is not yet in the training data of models – but it could be, if legal and economic barriers are overcome.
Large language models (LLMs) have historically been trained on vast swaths of internet text, often scraped without permission.
Recently, however, there is a notable shift toward formal content licensing for AI training data, especially in the U.S. and EU.
Major publishers, news firms, and other content owners are increasingly striking deals with AI developers, or pursuing legal action in absence of a deal, to ensure their content is used only with permission and compensation.
In this research article, we examine how likely it is that LLM developers will license content (versus unlicensed scraping), what factors are driving this shift, how much high-quality content remains untapped by AI, and who the key stakeholders are in the emerging content licensing industry.
‘These models need to be fed’
LLM developers are increasingly likely to license content rather than rely on non-permissioned scraping.
Over 2023–2025, multiple big deals have been announced in both the U.S. and Europe, signaling a trend.
For example, OpenAI has signed agreements with leading news publishers, including the Financial Times (UK) and News Corp (publisher of Wall Street Journal and others), to supply up-to-date news content and archives for ChatGPT and model training.
In March 2024, OpenAI also signed deals with France’s Le Monde and Spain’s Prisa Media (owner of El País and other outlets), allowing those publishers’ articles to be used in training and to surface as summarised, attributed ‘answers’ in ChatGPT.
Thomson Reuters’ CEO has mentioned “a number” of licensing deals to let AI companies train on Reuters news wire content, noting that “these models need to be fed… by the highest-quality, independent fact-based content”.
Such deals indicate that LLM creators are willing to pay for reliable data feeds rather than risk using copyrighted text without consent.
Legal and public pressure
Legal and public pressure are pushing AI firms toward licensing.
In late 2023, The New York Times filed a high-profile lawsuit against OpenAI (and its partner Microsoft) after negotiations broke down, alleging that OpenAI “taught ChatGPT to ignore journalists’ copyrights” and seeking damages and even the destruction of models trained on NYT content.
This case highlighted the legal risk of using copyrighted material without a license in the U.S., where the question hinges on how courts interpret fair use.
Similar lawsuits by groups of publishers have emerged. For instance, in early 2025 a coalition of 14 large U.S. news and magazine publishers (members of the News/Media Alliance) sued AI startup Cohere for “widespread unauthorised use” of thousands of their articles in training data.
Notably, some of these same publishers (e.g. The Guardian and The Atlantic) have licensed content to other AI platforms, underscoring that developers who don’t strike deals may face litigation.
In Europe, publishers have also taken action. Italian and Indian news firms have pursued legal remedies against OpenAI’s scraping, while others have chosen partnership, e.g. French news agency AFP signed a multi-year deal with AI startup Mistral in 2025 to provide its newswire content to the model “to strengthen the accuracy and relevance” of AI outputs.
Overall, the climate is increasingly favoring “deal or no deal”: AI developers who want access to high-quality, up-to-date data are finding it prudent to negotiate licenses, and those who do not are encountering lawsuits or public criticism.
Authoritative content is valued the most
Market trends and model performance needs also make licensing likely. As LLM development moves from academic experimentation to competitive commercial deployment, companies need premium data to differentiate and continually build their models.
Research shows that AI developers “favor high-quality content” from reputable publishers when assembling training datasets.
“Because AI models cannot effectively train themselves on their own output, known as synthetic data, they constantly require the new training data to evolve and maintain integrity,” says Stephen Conmy, content director with Contenseo.
“As a result, there is now a massive demand [and market] for quality AI training content and for smart ways to capture more data and feed it to new applications.”
In fact, an analysis by Ziff Davis found that as training data is curated more selectively, the share of content from major “premium” media domains rises dramatically, from well under 1% in raw web scrapes to over 12% in a curated dataset like OpenWebText2 (see table below).
This shows that the most authoritative content (news sites, scholarly sources, etc.) is disproportionately valued in making better AI models.
Because much of that high-value content is copyrighted (and often paywalled), licensing agreements are a logical solution to obtain it legally. In short, both the carrot (better data leading to better models) and the stick (lawsuits and laws) are motivating AI companies to formalise content access.
Table above: Share of content from top news/media publishers in various training datasets. Curated sets (OpenWebText, etc.) include a much higher fraction of high-authority publisher content than unfiltered web crawls. This indicates LLM developers deliberately seek out quality content, which often requires licensing.
Finally, leading AI companies themselves are signaling a shift to licensed data. OpenAI has publicly stated it is in discussions with “dozens of publishers” about licensing content for model training.
Google has likewise made deals, for example, a $60 million/year agreement with Reddit (signed Feb 2024) grants Google’s AI access to Reddit’s vast user-generated content.
These moves suggest that paying for data is becoming a norm.
Even open-source model communities are grappling with this trend. Early LLMs were trained on scraped datasets like The Pile, but now some open data sources are being withdrawn or restricted (e.g. Reddit’s API, Stack Overflow’s data).
Taken together, it is highly likely that future large LLM projects will rely far more on licensed corpora than on indiscriminate web scraping.
The unlicensed route is increasingly fraught with legal risk, and as the “easy” public data gets used up, model builders will turn to the remaining rich troves of content that require permission to access.
This wave of litigation exerts pressure
Several legal, economic, and industry factors are driving the move toward content licensing for AI:
Copyright law and lawsuits (U.S.): In the U.S., the legality of using copyrighted text for AI training without consent is unsettled, hinging on fair use. Rightsholders are aggressively testing this in court. Alongside the NY Times case, multiple class-action lawsuits by authors (organised by the Authors Guild and others) target OpenAI, Meta, and Google for training on copyrighted books. Authors argue that using books without a license isn’t fair use and are seeking to “clarify that books need to be licensed when used to train AI”.
Similarly, news publishers in the U.S. (News/Media Alliance members) accuse AI firms of copying articles (even paywalled content) without permission. This wave of litigation exerts pressure on AI developers to avoid unauthorised data use. Even if fair use might eventually protect some training uses, the cost and uncertainty of litigation is a strong deterrent. Thus, legal risk is pushing companies toward the safer path of licensing agreements, essentially “contracts not to sue” that grant needed permissions.
Regulatory environment (EU): The EU has established a stricter framework around text and data mining (TDM) and is considering new AI regulations that bolster licensing. Under Article 4 of the EU’s 2019 DSM Copyright Directive, rights holders can opt out of allowing their works to be mined for commercial AI training (for example by a machine-readable notice). If they opt out, AI developers must obtain a license or else infringe copyright. This opt-out mechanism effectively encourages licensing: many European publishers have indeed added notices reserving rights for AI/TDM on their websites. Furthermore, the proposed EU AI Act may soon require transparency on training data and proof of copyright compliance as a condition for deploying models in Europe. A recent draft would oblige AI providers to “ensure the training… is carried out in compliance with Union law on copyright…with regards to Article 4(3)” (the opt-out rule). If adopted, this will “significantly strengthen the position of creators and rights holders to prevent or license the use of their works” for AI training. In practice, that means AI companies operating in the EU would have to filter out any content that was opted-out or else secure a license, a strong legal driver for formal agreements. Even ahead of new laws, European media have leverage through national regulations (e.g. France’s neighboring right for news snippets) and a general legal climate favoring creators. The result is that AI firms find it necessary to negotiate with EU content owners (as seen by the flurry of deals with European publishers in 2023–24) to ensure compliance and avoid regulatory roadblocks.
Economic incentives: High-quality content is valuable fuel for AI, and content owners now realise this value. Publishers see an opportunity for new revenue streams by licensing archives and data to AI developers. The deals announced so far often involve substantial payments. For instance, the Google/Reddit deal at ~$60 million per year demonstrates the market value of a major platform’s user content. Publishers that license content can command significant fees or integrations; Reuters’ CEO noted in March 2024 that “a market price [is] evolving” for training-quality data. For the AI companies, these costs are weighed against the benefit of having exclusive or reliable access to data that can improve their models (and against the potential cost of litigation or bad PR if they don’t license). As training large models already costs tens or hundreds of millions of dollars in compute, the relative cost of data licenses is small in context, and paying for premium data can be seen as a necessary investment. There’s also a competitive angle: if one AI provider secures exclusive rights to a valuable dataset, it could gain an edge over rivals. This dynamic pushes the industry toward a data licensing economy, where content becomes a traded commodity for AI. Indeed, some observers predict a “boom in data licensing deals” for AI as companies race to secure scarce quality data. The corollary for content owners is the emergence of a “get paid or get scraped” ultimatum; and many prefer to get paid.
Industry relations and ethics: The tech and media industries are trying to establish a new equilibrium. Media companies have loudly protested that unbridled AI data mining “undercut[s] our businesses while using our own content to compete with us”, as a News/Media Alliance representative put it. In response, AI firms have begun outreach to assuage these concerns. Licensing deals are often bundled with promises of attribution or traffic, framed as partnerships rather than just transactions. For example, deals like OpenAI’s with Le Monde and FT include features where ChatGPT will cite and link to original articles, ostensibly driving readers back to publishers. This kind of “AI meets publisher” collaboration helps soothe industry fears and demonstrates a more ethical, consent-based model of AI development. There are also reputational stakes: as public awareness grows, AI companies risk backlash if seen as exploiting creators. Adopting a licensing approach can be part of responsible AI practices, showing respect for IP and artist rights. Even within AI firms, there’s recognition that cooperating with content creators is vital for long-term sustainability (e.g. OpenAI’s content policy shifts to allow site owners to opt out of its crawler). So, industry norms are evolving. What was common scraping practice a few years ago is now increasingly viewed as improper unless sanctioned. Tech companies that lead in embracing licensing may influence others and potentially help stave off harsher regulation by demonstrating goodwill.
Data scarcity and quality needs: Another factor is the sheer demand for more training data and the fact that easy pickings from the open web might diminish over time. Analysts warn that AI companies could“exhaust publicly available data for LLM training at the turn of the decade”, given the current exponential growth in model size and data requirements. Once the freely accessible text is mostly tapped, previously unused sources must be tapped, and many of those are behind licensing barriers (books, archives, etc.). Forward-looking AI developers are preparing for this by establishing content pipelines via licenses. Additionally, as models improve, simply adding more data is not as effective as adding better data. This incentivises curation of high-quality, diverse datasets, often requiring access to professionally produced content (journalism, literature, scientific writing) that isn’t freely available. All these strategic needs reinforce the move to licensing as the path to obtain reliable, rich training data in the long run.
A combination of legal pressure (copyright enforcement), regulatory trends, financial calculus, industry norms, and data needs is driving the shift. Especially in the U.S. and EU, where publishers and creators have strong voice, these factors converge to make licensed content usage the prudent course for LLM development going forward.
We are witnessing the foundation of a new content licensing industry for AI, where deals between AI firms and content owners become commonplace.
How much content remains untapped?
Despite the billions of words already ingested by GPT-style models, there remains a vast amount of high-quality content not yet used (or only minimally used) in LLM training. This includes whole genres of copyrighted text and data that have been either off-limits or technically out of reach.
Below, we estimate how much valuable content is still untapped, identify major categories of such content, and explain why each remains underused by AI.
It is challenging to quantify precisely how much new text data remains unused, but several indicators show that the reserve of high-quality content is enormous.
Consider books: an infamous bootleg dataset called Books3 contained about 183,000 pirated e-books, and was used (without approval) in some AI research models.
Even if some AI developers covertly scraped more books, it’s clear that a huge volume of literature remains untouched by current LLMs.
Similarly, in scholarly and professional literature, only a fraction has been used. There are over 50–100 million academic journal articles in existence (with 3+ million new articles published each year) but LLM training data has largely come from openly available subsets (such as arXiv preprints or open access journals).
The bulk of peer-reviewed articles behind paywalls (spanning medicine, science, humanities, law) likely remain absent or under-represented in training corpora.
Major scientific publishers have not broadly opened their vaults to AI (barring specific deals), so this content is still “dark matter” to models.
Much more archives remain untouched
Another reservoir is news and magazine archives. While models like GPT-4 have read some online news, they have not necessarily been trained on full historical archives of leading newspapers or magazines (which often require expensive licenses or are not digitised for open web).
For example, the New York Times has archives going back to the 1850s; the Times (London) to the 1780s; countless local newspapers have century-plus archives; a rich record of human events and language.
Much of this is high-quality edited content that is not freely accessible on the web (often behind archive paywalls or library systems) and thus not in common datasets.
User-generated content also contains untapped pockets
Although forums like Reddit and Wikipedia were heavily used in training, other platforms were not. Reddit itself moved to restrict free access in 2023, so newer Reddit content may be unused unless licensed (hence Google’s paid deal to get it).
Twitter (now X) similarly cut off free API access, meaning the vast corpus of tweets is largely inaccessible to new AI models unless through a paid arrangement.
Private or semi-private communities (Facebook posts, private subforums, etc.) are of course mostly off-limits.
Even government and legal documents, many of which are public domain, can be missed because they aren’t compiled in convenient machine-readable form (for example, court transcripts, legislative histories, or government archives that are not web-indexed).
What remains unused is not so much random text as it is valuable text locked behind barriers.
In our next research piece we will examine just how much valuable content and data has not yet been used by LLMs and the reasons why it hasn’t yet been used.
The rapid advancement of generative artificial intelligence has created a significant demand for high-quality training data.
AI companies like OpenAI, Google, Meta, and others require vast amounts of content to train their large language models and other generative systems.
However, much of this valuable content is protected by copyright, necessitating licensing arrangements between AI developers and content owners.
Here, we examine the current landscape of content licensing in the generative AI industry, focusing on the types of licensing deals, financial terms, and evolving best practices.
Types of licensing arrangements
Generative AI companies employ several different licensing models when securing content from copyright holders.
Each approach offers different advantages and limitations for both parties involved.
1. Direct one-to-one licensing
The most common approach involves direct negotiations between an AI company and a content owner, resulting in custom licensing agreements.
Direct licensing offers content owners maximum control and potentially higher compensation, but requires significant negotiating resources.
This model involves a third-party aggregator that collects and licenses content from multiple sources.
For content owners, this reduces administrative burden while still providing legal protection and compensation.
AI companies benefit from streamlined access to diverse content through a single relationship.
Examples include licensing arrangements with stock content providers like Shutterstock and Getty Images, which aggregate visual content from multiple creators.
3. Collective licensing
Though less common in the US currently, collective licensing involves organisations that represent groups of content owners and negotiate on their behalf.
This model provides broader representation and potentially more equitable compensation for smaller content owners. It reduces transaction costs for both sides but may result in lower individual payouts.
4. Emerging alternative models
Several innovative licensing models are beginning to emerge:
Revenue-sharing platforms such as Perplexity’s Publisher Program, offering content owners a percentage of ad revenue generated when their content is cited in AI responses.
Financial terms and compensation models
The financial aspects of AI content licensing deals vary widely based on content type, volume, exclusivity, and the parties involved.
1. Monetary compensation structures
Flat fee arrangements
Many licensing deals involve fixed upfront payments:
OpenAI’s 5-year deal with News Corp: Over $250 million, including both cash and credits for technology use.
Microsoft’s licensing deal with Informa: Over $10 million.
OpenAI offers between $1-5 million per corpus for some deals.
Some news organisations receive offers starting at $1 million.
Exclusive: Content is licensed exclusively to one AI company, usually commanding higher fees but limiting future options.
2. IP protection and attribution
Copyright preservation
“Agreements typically maintain the original copyright holder’s ownership rights while granting specific usage permissions,” says Stephen Conmy, director with Contenseo.
Attribution requirements
Many deals include:
Citation mechanisms: Provisions for how AI outputs should cite the original content.
Linking requirements: Specifications for how links to original content should appear.
Branding guidelines: Rules for displaying the content owner’s brand in AI outputs.
3. Quality control and oversight
Content review rights
Publishers often retain:
Rights to review AI-generated content that uses their material.
Approval processes for how their content is represented.
Usage monitoring
Agreements typically include:
Detailed analytics requirements: Provisions for tracking how content is used.
Transparency obligations: Requirements for AI companies to provide visibility into content usage.
4. Duration and termination
Licensing agreements generally specify:
Term duration: Often multi-year agreements (e.g., OpenAI’s 5-year deal with News Corp, Shutterstock’s 6-year deal with OpenAI).
Renewal provisions: Terms for extending the agreement.
Termination clauses: Conditions under which either party can end the agreement.
Case studies of notable licensing deals
OpenAI’s publisher partnerships
OpenAI has been particularly active in securing content licensing deals:
News Corp Agreement: A 5-year deal worth over $250 million that includes content from The Wall Street Journal, New York Post, and other properties. The deal includes both cash payments and credits for OpenAI technology.
Axel Springer Partnership: Worth over $25 million, providing access to content from outlets like Politico and Business Insider. The agreement includes enhanced visibility for Axel Springer’s content in ChatGPT responses.
Associated Press Deal: Estimated between $1-5 million annually, granting OpenAI access to AP’s news archive dating back to 1985.
TIME Partnership: A strategic agreement that includes fixed fees, performance-based compensation, and technology integration benefits.
Google’s content licensing
Google has established several significant agreements:
Reddit deal: A multi-year agreement worth approximately $60 million annually, with a total contract value of about $203 million. The deal provides Google with access to Reddit’s user-generated content for AI training.
Stack overflow partnership: An agreement that includes licensing Stack Overflow’s technical Q&A content and integrating with Google Cloud.
Stock media licensing
Visual content providers have been key partners for AI companies:
Shutterstock’s multiple agreements: Deals with OpenAI (a 6-year agreement), Meta, Amazon, and Apple valued between $25-50 million each. These arrangements have helped Shutterstock’s Data Distribution segment grow from $15.9 million in 2021 to approximately $137 million in 2023.
Getty images partnerships: Licensing arrangements with multiple AI companies including Runway, with terms that include strict attribution requirements and usage limitations.
Challenges and considerations in content licensing
For content owners
Content owners face several challenges when licensing their material for AI training:
Valuation difficulties: Determining appropriate compensation for content used in AI training is complex.
Negotiating leverage imbalance: Smaller content owners often have limited bargaining power compared to major AI companies.
Long-term implications: The potential impact of current licensing decisions on future revenue opportunities remains uncertain.
Tracking compliance: Ensuring AI companies adhere to licensing terms can be difficult without proper monitoring mechanisms.
For AI companies
AI developers face their own set of challenges:
Escalating costs: The price of high-quality training data continues to rise, creating financial pressure.
Legal uncertainty: The evolving legal landscape around copyright and AI training creates risk.
Data quality verification: Ensuring licensed content meets quality standards for effective model training.
Managing multiple agreements: Coordination of various licensing deals with different terms can become administratively complex.
Future trends in AI content licensing
Several emerging trends are likely to shape the future of AI content licensing:
Standardisation efforts: Industry initiatives to create more standardized licensing frameworks to reduce negotiation complexity.
Collective licensing growth: Increasing adoption of collective licensing models to improve efficiency and equity.
Dynamic pricing models: Evolution toward more responsive pricing based on actual content usage rather than flat fees.
Regulatory influence: Growing impact of legislation and regulations on licensing practices, potentially mandating fairer terms.
Market expansion: The overall market for AI training data is projected to grow from approximately $2.5 billion currently to nearly $30 billion over the next decade.
A rapidly evolving ecosystem
The licensing of content by generative AI firms represents a rapidly evolving ecosystem with significant financial implications for both AI developers and content owners.
As the industry matures, we’re seeing increasing standardisation in deal structures, though considerable variation remains in compensation levels and specific terms.
The most successful licensing arrangements balance fair compensation for content creators with sustainable economics for AI companies, while maintaining appropriate content usage controls and attribution requirements.
As the generative AI sector continues to grow, we can expect further refinement of licensing practices and potentially new models that better align the interests of all stakeholders.
For content owners considering licensing their material for AI training, careful attention to both financial terms and usage rights is essential.
For AI companies, establishing clear, fair licensing practices not only reduces legal risk but also helps ensure access to the high-quality content needed for effective model training.
As regulatory frameworks evolve and market dynamics continue to shift, both sides of these licensing arrangements will need to maintain flexibility while working toward more standardised approaches that can scale efficiently with the growing demand for AI training data.
As the generative AI sector continues to grow, we can expect further refinement of licensing practices and potentially new models.
A US federal appeals court has ruled that a machine cannot be named as the author of a work, when a human submits it to the U.S. Copyright Office.
On March 18, 2025, a D.C. federal appeals court ruled in Thaler v. Perlmutterthat the Copyright Office was right to reject Dr. Stephen Thaler’s claim for an AI-generated image called ‘A Recent Entrance to Paradise’ (pictured).
Thaler, a computer scientist, built a generative AI called the ‘Creativity Machine’, which produced the image on its own.
Thaler then tried to register the copyright of the image. He listed the ‘Creativity Machine’ as the sole author and himself as the owner of the work.
Thaler appealed, but the federal court upheld the decision.
“The Creativity Machine cannot be the recognised author of a copyrighted work because the Copyright Act of 1976 requires all eligible work to be authored in the first instance by a human being,” said Circuit Judge Patricia A. Millett.
AI can help humans, but not be the sole creator
There are, however, nuances when it comes to the use of AI and creating ‘work’.
In January 2025, a U.S. Copyright Office report declared that creators and artists can copyright works made with the help of AI.
As AI and tech giants push for unrestricted access to limitless training data, a battle is brewing between the world’s top creators and the ever-hungry language models that thrive on human creativity.
Two of the planet’s most powerful AI developers, OpenAI and Google, recently submitted letters to the White House Office of Science and Technology, responding to its request for input in the development of an AI Action Plan.
The two multibillion-dollar corporations are urging the Trump administration to grant them the freedom to disregard copyright laws, allowing them to use any material they choose to train their language models.
OpenAI argues that the U.S. will lose its competitive advantage in the race to become the world’s AI leader if it strictly adheres to intellectual property frameworks.
Major Hollywood figures were quick to push back on these calls. Nearly 500 actors and creatives signed a letter arguing that “America’s global AI leadership must not come at the expense of our essential creative industries.
It is clear that Google (valued at $2 trillion) and OpenAI (valued at over $157 billion) are arguing for a special government exemption so they can freely exploit America’s creative and knowledge industries
“AI companies are asking to undermine this economic and cultural strength by weakening copyright protections for the films, television series, artworks, writing, music, and voices used to train AI models at the core of multi-billion dollar corporate valuations.”
Read the full letter below.
We, the members of America’s entertainment industry — representing an intersection of cinematographers, directors, producers, actors, writers, studios, production companies, musicians, composers, costume, sound & production designers, editors, gaffers, union and Academy Members, and other industrious, creative content professionals – submit this unified statement in response to the Administration’s request for input on the AI Action Plan.
America didn’t become a global cultural powerhouse by accident. Our success stems directly from our fundamental respect for IP and copyright that rewards creative risk-taking by talented and hardworking Americans.
We firmly believe that America’s global AI leadership must not come at the expense of our essential creative industries. America’s arts and entertainment industry supports over 2.3 million American jobs with over $229 billion in wages annually, while providing the foundation for American democratic influence and soft power abroad. But AI companies are asking to undermine this economic and cultural strength by weakening copyright protections for the films, television series, artworks, writing, music, and voices used to train AI models at the core of multi-billion dollar corporate valuations.
Make no mistake: this issue goes well beyond the entertainment industry, as the right to train AI on all copyright-protected content impacts all of America’s knowledge industries. When tech and AI companies demand unfettered access to all data and information, they’re not just threatening movies, books, and music, but the work of all writers, publishers, photographers, scientists, architects, engineers, designers, doctors, software developers, and all other professionals who work with computers and generate intellectual property. These professions are the core of how we discover, learn, and share knowledge as a society and as a nation. This issue is not just about AI leadership or about economics and individual rights, but about America’s continued leadership in creating and owning valuable intellectual property in every field.
For nearly 250 years, U.S. copyright law has balanced creator’s rights with the needs of the public, creating the world’s most vibrant creative economy.
It is clear that Google (valued at $2 trillion) and OpenAI (valued at over $157 billion) are arguing for a special government exemption so they can freely exploit America’s creative and knowledge industries, despite their substantial revenues and available funds. There is no reason to weaken or eliminate the copyright protections that have helped America flourish. Not when AI companies can use our copyrighted material by simply doing what the law requires: negotiating appropriate licenses with copyright holders — just as every other industry does. Access to America’s creative catalog of films, writing, video content, and music is not a matter of national security. They do not require a government-mandated exemption from existing U.S. copyright law.
America didn’t become a global cultural powerhouse by accident. Our success stems directly from our fundamental respect for IP and copyright that rewards creative risk-taking by talented and hardworking Americans from every state and territory. For nearly 250 years, U.S. copyright law has balanced creator’s rights with the needs of the public, creating the world’s most vibrant creative economy. We recommend that the American AI Action Plan uphold existing copyright frameworks to maintain the strength of America’s creative and knowledge industries, as well as American cultural influence abroad.
Created by Humans founders Jen Singerman (left), Trip Adler (centre), and Edward Igushev (right).
A coalition of American novelists and authors is banding together to safeguard their intellectual property and prevent AI companies from pilfering their works to train language models, thereby preserving the copyright and integrity of their creative output.
“Authors must come together now and negotiate with AI companies as a block,” said novelist Doug Preston in an interview with Fortune.com.
Preston says he has published at least 40 books, all of which he has discovered have been used by AI firms to train their models without his consent.
Violation
This violation of his copyright led him to believe that authors and AI developers need some sort of licensing system that protects both parties from copyright infringement.
Created by Humans was launched in June 2024 and acts as a licensing service for authors. It is a licensing bridge between authors and AI developers.
“Authors can sign up, claim their works, and note the licenses they’d like to grant to AI developers,” said Jen Singerman, co-founder and COO of Created by Humans.
Why this matters
Creators of all kinds are increasingly frustrated with how their work is being used by AI firms without their permission.
According to CbH, over 100 authors have signed up to the service (as of January 2025).
The Authors Guild, a professional group representing over 15,000 authors, including its former president Preston, has partnered with Created by Humans since October 2024.
“We urgently need to give control back to authors and their publishers, and licensing is the means to accomplish that going forward,” said Mary Rasenberger, Authors Guild’s CEO.
Money for work
Preston notes that AI licensing could be a lucrative opportunity for authors, as AI companies rely on books to train their large language models, and require repeated ingestion of new texts with each model iteration.
The CEO of Created by Humans, Trip Adler, claims CbH is currently AI licensing a book every five minutes.
Industry pricing models
The big question is, how much can authors charge AI companies when licensing their work? There are four pricing models currently in place.
Flat fee – A one-time payment for access to an author’s catalog (e.g., $50,000 – $500,000+ depending on content volume and uniqueness).
Per-word rate – AI firms sometimes pay per million words (e.g., $1,000 – $10,000 per million words, based on exclusivity and quality).
Revenue share – Authors receive a percentage of the revenue AI firms generate using their content.
Tiered pricing – Authors charge different rates based on usage (e.g., research vs. commercial use).
In November 2024, the Authors Guild disclosed that HarperCollins, a leading publisher under NewsCorp, had entered into an agreement with Microsoft, allowing the tech giant to use select HarperCollins nonfiction titles to train its AI models. The deal priced each title at $5,000 for a three-year period of training rights.
Fundraising
Created by Humans has raised a total of $11 million in funding, including a $5.5 million seed round led by Giant Ventures.
Once you write something, whether it’s a piece of journalism, a guide or a book, you become the copyright owner of that work.
So why does this matter? It matters because you are now the owner of a creative asset. And you can license that work to others.
What is copyright?
‘Very simply, copyright is the protection of an original work,’ says Stephen Conmy, founder of Contenseo, ‘and content licensing is when someone gets permission to re-use that work. Usually there is a fee involved.’
Copyright law grants creators the exclusive right to reproduce, distribute, and display their original work, such as articles, photographs, music, and other works.
As more and more writers and creators are no longer producing work exclusively for one publisher, the opportunity to license their work has increased significantly.
The wide, rich world of content licensing
Making your work available to license gives you access to more revenue, a wider audience, the protection of your intellectual property, and the chance to collaborate with potential new clients.
‘Content licensing allows creators to share their work with a broader audience, increasing visibility and reach,’ says Conmy.
Licensing agreements can help protect intellectual property rights, ensuring that creators receive fair compensation for their work.
Content licensing creates more money for writers
Quality writing that has mass appeal is the type of writing that gets licensed the most.
Most great commercial writing is, at its heart, investigative journalism. “By investigative journalism, I don’t just mean crime reporting,” says Conmy. “Investigative journalism involves a journalist or a group of journalists working on a story to extract as many verified details as possible. It’s about research, interviewing people, and analysing any evidence and content around the story.”
Copyright and licensing agreements can protect a writer’s investigative work, ensuring that their research and findings are not misused or exploited, particularly by AI companies that may just take the work to train their models without the writer’s permission.
Along with protecting their copyrights and their research, by making their work available to license, journalists can generate revenue through syndication.
Licensing agreements can also help journalists reach a broader audience, increasing the impact of their work.
And then there is the issue of protecting sources. Journalists can use copyright and licensing agreements to protect the identities of their sources, ensuring their safety and security.
How to start licensing your content
“Contenseo is an abbreviation of ‘Content Licensing Software’, we help creators license their content,” says Conmy.
All you have to do is open a Contenseo account – the ‘Licensing as a Service’ model. You will then be able to upload your work. All you need is the URL of the webpage on which the work was originally published.
When you have registered your work, you can then get access to Contenseo’s licensing agreements to protect your intellectual property and ensure fair compensation for your work.
And remember, Contenseo offers you the opportunity to collaborate with other publishers and creators, helping to build a community and promote a culture of creativity and innovation.
In short, by becoming a member of Contenseo, you will learn how to use copyright and content licensing, protect your intellectual property, generate revenue, and share your work with a wider audience.
© 2024 Contenseo. All rights reserved.