AI royalties finally available for small and medium publishers[19459005]

The RAG (retrieval augmented generation) kicks in when the LLM is prompted by a user or system. It then pulls relevant content from different sources to provide the best answer.

Consider Dow Jones’ Factiva which operates an AI-licensed enterprise content marketplace. It has quietly expanded its licensing agreements for generative AI use over the last two-years, with around 5,000 publishers out of its 30,000 strong publisher network opting in. Emma O’Brian (General Manager at Factiva) was hired to this role in August. She said that it was a long-term undertaking, but it has helped the unit become an effective negotiator for AI licensing rights to unlock royalties on a usage-based basis for small and medium-sized publishers.

Dow Jones also houses The Wall Street Journal, both under News Corp parent company. O’Brian said this was important to know when negotiating the best terms for other publishers. She said, “We wouldn’t sign a [AI licensing] contract we wouldn’t sign ourselves.”

This includes having a conversation with enterprise clients about how they plan to use content after it has been RAG’ed. This is to ensure that there will be no future risk of leakage of data, and to determine how many people are going to be using it.

A client might need to specify whether they plan to use content for the internal workforce or to generate a report — an entirely different use case. Once publishers have been onboarded, the Factiva team will then work with the tech department or data ingest department of its corporate clients to see how the system works in practice and ensure that there aren’t any risks of data leakage. O’Brian said that the team will check in periodically to make sure they are adhering to their contract.

Factiva uses a usage API that ensures that usage can be attributed and that its publishers will receive accurate payments (per usage per RAG query), according to O’Brian.

O’Brian’s priority is to ensure that specialist, niche B2B companies get royalties and not just the bigger ones. Factiva should also be providing the type of specialist content required for specific industries, rather than aiming at a more general use.

O’Brian said, “I don’t call it a’moat’, but I think that when you compare the gold ore we have in trusted and reliable content that businesses can RAG against versus the garbage in garbage out [noisy scraped data] models to broadly scraping, that’s an enormous opportunity.” “And that is something that cannot be scraped. It has to be ring-fenced.”

Eye on the future: Factiva as negotiator for consumer LLMs.

Factiva, with its thousands of publishers licensed and direct relationships with them across its network is uniquely positioned to be a collective licensor – aggregating rights to negotiate B2C agreements with major LLM platforms. This could turn scattered archives into revenue streams for publishers that would never have been able to land these talks alone. O’Brian said that this is a route they intend to actively pursue.

To clarify, it is in a kicking-the tires mode today. The real action today remains firmly on B2B – prying open usage based royalties for publishing via enterprise clients’ LLMs. It’s not blinded by the potential negotiating strength it could yield if its network gave it the go-ahead to negotiate with consumer-facing LLMs.

The existential question is, “How are our publishers managing their relationships with LLMs directly ?,'”,” said O’Brian. “That’s an important consumer question and we hope to be able to play a larger role in the future.” We want to sit at the table with publishers and have conversations about how we can help.

The growth of enterprise LLMs has opened up new AI licensing opportunities, including for consumer titles like The Economist or The Financial Times. Other collective licensing and rights organisations are also moving towards the same direction, carving out a promising stream of revenue for long-tail publishers.

Opens AI royalties to long-tail publishers.

The growing demand for quality content for private LLMs and the increasing popularity of small language models are a source of hope for long-tail publishers who are usually left out of larger LLM licensing discussions. This was stressed by Tom West, the CEO of Publishers’ Licensing Services, a non-profit collective management organisation representing U.K. Publishers in collective licensing.

Tom West said that small LLMs are specializing in specific industries or topics, such as horticulture. They are therefore on the lookout for quality specialist content to be published by niche publishers. Because enterprises value provenance and compliance, vetted journalism, and evergreen explainers, become premium inputs with clear attribution, and tighter controls than consumer AI.

As companies increasingly use AI for research, data analysis, summarizing articles, and admin work, it is important to ensure that the material fed into these tools is properly licensed. According to the Copyright Clearance Center (CCC) study 2025 Content Usage Trendsonly 62 percent employees are aware that their company’s copyright policies also cover using third-party material with AI.

Smaller publishing groups can tap into this demand by partnering with vertical aggregators, while larger groups are able to strike direct deals that include data. It’s not flashy but it’s durable – and it turns editorial back catalogues into recurring enterprise revenue.

West stated that it is impossible to calculate how much revenue publishers will make in this way at this time, due to the fact that it is still early and they have not yet created a licensing template. It pays out more than PS50 million ($66.5million) per year to publishers through collective licensing, and half of this comes from corporate licensing. “I am really excited about the potential to curate content for smaller organizations, package it together and then license to start-ups. [creating SLMs]He said that the market was untapped.

PLS is owned by U.K. Trade bodies, including the Professional Publishers Association. It works with the Copyright Licensing Agency to handle copying licenses, and returns money to publishers.

The CCC in the U.S. has a similar roadmap. Lauren Tullock is the vp and managing Director at the CCC. She said that it has adapted copyright licensing agreements for AI usage. This includes private LLMs of corporate enterprises. The licenses are based on usage data collected by surveys and are opt-in only for rights holders. CCC has thousands and hundreds of publishers for AI rights. The CCC will calculate usage and pay out royalties every six months.

She said, “It is a very good thing for small and mid-sized publishers whose content is very valuable for certain sectors. It’s difficult for clients to manage direct arrangements with each one of them.”