How AI Publisher Licensing Deals Shape Which Brands Get Cited

AI publisher licensing deals shape which publishers AI systems can quote and attribute, which means the publishers writing about you increasingly matter as much as your own pages. Not content quality alone. Not domain authority.

The pace tells you how fast the stakes are rising. Publicly announced agreements have gone from a handful in 2023 to dozens by 2026, and the newer ones put far more weight on attribution and real-time display than the 2023 wave did. Several of the largest 2025 and 2026 deals still include training rights, so read each one rather than assuming the category has moved.

The Reddit deal is the most publicly reported example. Each agreement controls what language enters a model's training data and what surfaces when buyers query AI tools today.

Key Takeaways

• Two distinct mechanisms govern this: foundational training data and real-time retrieval. Each has a different timeframe and requires a different approach.
• Most brands conflate them, and optimize for the faster one.
• Between 40% and 60% of AI-cited sources rotate month to month. Brand citation isn't a ranking you earn once.
• A five-step audit gives you a measurable baseline for where your brand stands today.

What are AI publisher licensing deals?

AI publisher licensing deals are commercial contracts between AI companies and content owners, granting AI labs the right to use published content for model training or real-time retrieval.

Licensing agreements specify how an AI company may use a publisher's content. Depending on the contract, that can include model training, retrieval for current answers, or both. Public announcements may also describe attribution and linking arrangements. A licensing agreement shouldn't be assumed to include training rights unless those rights are stated.

The pace of these deals has accelerated since 2023, driven by two pressures:

• AI labs needed to expand training datasets beyond what open web crawls could provide.
• Publishers threatened legal action after discovering their content had been used without permission or compensation.

The result is a commercial layer sitting between published content and the AI systems your customers use to make purchase decisions.

What do those contracts actually buy?

The answer depends entirely on whether the deal covers training data or retrieval access. These are fundamentally different, and they affect brand visibility in different ways.

A training data deal shapes what the model has learned before it answers any question. Call it long-term memory. Brands appearing consistently in that training material become part of the model's baseline understanding of a category, which is what it will say without looking anything up.

A retrieval deal works differently. Content is accessed in real time while the model answers a specific query, which the industry calls retrieval-augmented generation. The content doesn't change what the model knows permanently. It shapes that one response.

Not all these deals are the same. The early wave, roughly 2023 to 2024, was mostly about training data: AI companies paying for historical archives to build better models.

The newer deals read differently, and the difference repays reading precisely. Washington Post's April 2025 agreement with OpenAI makes no mention of training at all: it covers summaries, quotes and links in ChatGPT. The Guardian's February 2025 deal reads the same way. Silence isn't the same as exclusion, and only Le Monde went further, explicitly ruling training out of its Perplexity deal.

Two caveats on reading it. Deal type is what the parties announced, and several announcements are silent on training rather than excluding it, which is a different thing. And the Wall Street Journal reported in July 2026 that the Google and Reddit agreement is ending soon, with Reddit having discussed shutting off Google's access while renewal talks continue, so one of the two largest published figures may change.

What the table doesn't show is what the money buys, and which brands benefit as a result. That's where most brand strategy goes wrong.

The Reddit deal does both. OpenAI gained Data API access for pulling current discussions during queries, and the right to train on Reddit's corpus. That's why Reddit is one of the few sources present in both layers at once, which is rarer than the volume of coverage suggests.

The News Corp deal is the clearest case of dual-platform positioning. One media portfolio holds an OpenAI deal worth roughly $250M over five years and a separate Meta deal worth up to $50M per year. That puts the Journal's reporting inside the retrieval layer of two major assistants at once, the kind of dual presence a single-platform deal can't buy.

For brand visibility, the practical difference is this:

• Training data shapes the model's unprompted recommendations.
• Retrieval data shapes its sourced answers.

Your strategy needs to address both, and most frameworks address one.

What is the difference between training data and RAG?

Training data is what the model carries into every conversation. RAG is what it goes looking for during one.

The difference shows the moment you ask a category question. Ask for the best project management software for a remote team, and named brands often appear with no cited source. That's foundational knowledge: those brands have been in enough high-authority content, over enough time, that the model associates them with the category by default.

Ask which project management tools launched in 2025 and the model will typically search and return citations. That's retrieval. It went looking because its foundational knowledge has a cut-off date.

The two require different actions:

• Foundational visibility comes from sustained presence in high-authority content over months and years, and compounds slowly.
• Retrieval visibility comes from being on the right platforms with the right content structure, improves quickly, and changes constantly.

Most brands chase retrieval visibility because it's faster to act on and easier to measure. Foundational visibility is the slower work, and the work competitors can't copy in a quarter.

Evertune, which sells base-model measurement, argues that no real-time optimization fully compensates for a weak training footprint. Treat that as a vendor position rather than published research, though its method of querying base models with search disabled is the right way to test the question.

What are the four signals that drive AI citations?

Four categories shape whether your brand appears in AI-generated answers. Most GEO frameworks cover three, and the fourth is the one that builds long-term, unprompted visibility.

Entity clarity

AI models organize knowledge around entities: named companies, people, products and concepts with consistent, verifiable attributes. A brand appearing across sources with a consistent name, description and category is easier for a model to represent accurately.

Entity clarity comes from schema markup, Wikipedia presence, consistent information across business directories, and persistent brand language across authoritative publications. Without it, a model may hold fragmented knowledge of your brand, or merge you with a competitor in the same category.

Content extractability

AI systems favor content that leads with the answer. A paragraph burying its key claim in the fourth sentence is harder to incorporate into a synthesized response than one opening with the point.

This isn't about bullet-pointing everything. It's about removing the narrative build-up that works for human readers and adds noise for a system trying to extract a usable answer.

Platform authority

AI models pull from specific sources: Reddit threads, YouTube transcripts, LinkedIn articles, peer-reviewed publications and licensed news. A brand participating genuinely in these environments builds presence across the platforms these systems draw on.

Reddit matters here not simply because OpenAI paid for the API. Real user discussions about your brand are the kind of third-party signal these models surface.

The weight of that signal is moving, though. Ahrefs' Brand Radar data shows ChatGPT's Reddit citations in the US falling 88% between March and late September 2026, while Reddit remains among its most-cited domains. Check how often Reddit appears in AI answers for your own topics before you build a plan around it.

Foundational model footprint

Most GEO frameworks leave this out. It builds through sustained presence in high-authority, training-eligible content: coverage in publications AI labs license, academic references, media mentions, and consistent brand association across years of authoritative text.

Building it takes 12 to 18 months. It's a program, not a campaign.

Why does traditional SEO fall short in AI search?

Traditional SEO is built on ranking logic. Content earns a position through signals an algorithm evaluates: backlinks, relevance, click-through rate, site health.

AI search has no ranking index. It has a generation process. The model doesn't sort your content against competitors and return the best match, it synthesizes a response from what it knows. The competition isn't for rank position, it's for inclusion.

The numbers support that, with a caveat about which numbers. Ahrefs put 37.1% of AI Overview citations in the top ten organic results in a March 2026 study, meaning most cited pages rank outside it. Ahrefs reports that share falling from roughly 76% eight months earlier, while warning its two datasets aren't directly comparable because it improved its parsing in between, so treat the direction as solid and any single percentage as provisional.

The measurement gap is sharper. Traditional SEO gives you rankings, traffic and click data. AI visibility gives you none by default: a brand can appear in thousands of AI responses a month and record zero corresponding traffic, because the user got a complete answer without clicking.

Then there's volatility. Semrush's AI Visibility Index, tracking 2,500 prompts across Google AI Mode and ChatGPT, found 40% to 60% of cited sources change month to month. Two platforms and a fixed prompt set, so treat it as a scale rather than a constant. AI visibility isn't a position you defend, it's a share of voice you manage.

Can brands without publisher deals still get cited?

Yes. The licensing deals shape the playing field. They don't close it.

Publisher agreements can give AI platforms access to licensed content, but brands can also appear through sources on the open web. A licensing agreement is neither a requirement for being cited nor a guarantee of inclusion.

Ahrefs' October 2026 research adds evidence about how often partner publications appear. In its separate Brand Radar citation dataset of US ChatGPT answers, OpenAI partner titles among ChatGPT's 200 most-cited domains increased their combined citation share from 7.5% in March to 15.8% in July, before settling at 12.5% in late September.

The Ahrefs figures describe publisher citations, not the visibility of individual brands mentioned in their articles. They also don't show that licensing agreements caused the increase.

For B2B teams, the practical implication is to consider relevant publisher coverage alongside your own content. Prioritize publications your buyers trust, then check whether those publications appear in AI answers about your category.

Future Publishing ran a campaign for Samsung using its own AI visibility tooling and reported 28% growth in AI citations from Future-sourced content over three months. Samsung is not party to any publicly announced content licensing deal with OpenAI or Google. Read the number as one vendor reporting on its own campaign rather than an independent measurement, and the mechanism still holds.

A smaller version of the same logic works for mid-sized brands. A B2B software company with a maintained LinkedIn presence, genuine participation in relevant communities, and a site structured for extraction can build meaningful AI visibility without a publishing deal. It takes longer than a campaign and compounds in a way a campaign never does.

How do you audit your AI visibility?

Run a five-step quarterly audit to establish where your brand stands, rather than optimizing from an assumed baseline. Most GEO content describes tactics without measuring the gap those tactics are meant to close.

Step 1: baseline your citation rate. Query each major platform with 10 to 15 category-level questions relevant to your product. Record whether your brand appears, how it's described, and whether the citation includes a source or appears to draw from foundational knowledge. Run each query three times in a fresh thread and average the results.

Step 2: audit your entity footprint. Check for consistent representation across Google Knowledge Panel, Wikipedia, Wikidata, Crunchbase, your LinkedIn company page, and your own schema markup. Inconsistencies in name, category or founding year are signals of entity ambiguity a model may carry into its answers.

Step 3: audit content extractability. Take your 10 highest-traffic pages. Does each paragraph lead with its conclusion? Is the key claim in the first sentence? Are headings descriptive? Score each page and find where the gap is largest.

Step 4: audit third-party coverage. Review mentions of your brand across relevant industry publications, news sites, Reddit, and YouTube over the past 90 days. Record the source, linked page, and context of each mention.

For priority publications, check whether they appear as sources in your repeated AI query tests. Note any publicly announced AI licensing partnerships, but prioritize relevance to your buyers and observed citation patterns when choosing where to pursue coverage.

Check business listings here too, with a caveat about how far the evidence stretches. Yext analyzed 6.8 million citations from location-context queries in July and August 2025, and found listings drove 42% against 44% for first-party sites. That study ran across retail, financial services, healthcare and food service. If you sell pure software, don't expect listings to carry that weight.

Step 5: measure monthly share of voice. Run the citation rate audit monthly, tracking two to three direct competitors on the same query set. Our guide to measuring AI share of voice sets out the method. The quarterly investment is four to six hours, and what it produces is something most brands don't have: a number.

FAQs

Where to start

Understand the difference between what AI models know and what they look up, build a strategy addressing both, and measure where you stand before optimizing anything.

Deals will multiply and their contents will turn over. The brands best insulated from that are those with foundational model knowledge already embedded, and broad presence across multiple retrieval sources. Building broad rather than deep is the more defensible position.

Want to know where your brand stands in AI search right now? Tenpoint Labs runs AI visibility audits for B2B and consumer brands. Get in touch to start with a baseline.