Every major publisher has signed an AI licensing deal. The press releases arrive with fanfare: strategic partnership, transformative collaboration, the future of content. The numbers, when you find them, tell a different story.
News Corp — the biggest deal in publishing — gets roughly $100 million a year from OpenAI and Meta combined. That sounds like real money until you remember News Corp's annual revenue is north of $10 billion. It's less than 1%.
People Inc's AI licensing revenue is harder to isolate. Their "licensing and other" line grew 26% year-over-year to $40.7 million in Q1 2026, but that includes Apple News+, syndication, and everything else that isn't advertising or subscriptions. The pure AI component — the OpenAI deal, the Microsoft arrangement, the Meta agreement — is likely $20 to $30 million annually. Against $1.8 billion in total revenue, that's a rounding error.
The Market That Isn't
David Skok, CEO of The Logic, made the structural argument in late 2025: as long as Google's search crawler and its AI-training crawler function as a single system, the licensing market is effectively dead. Publishers face a binary choice — allow AI training to maintain search visibility, or block everything and disappear. Most can't afford to disappear. So AI companies get training data for free from the largest search engine on earth, and the licensing deals that do get signed are, in Skok's framing, "public relations exercises" rather than price discovery.
The industry seems to agree. Digiday's survey of publishers found only 20% expect AI licensing to become a major revenue source. Nearly half say it will be a minor contribution at best. USA Today's CFO warned the revenue would be "lumpy or unpredictable."
The Deal Beneath the Deal
Here is what makes People Inc's OpenAI arrangement different from most publisher licensing deals: the check isn't the point.
The value chain hidden inside a licensing deal
Step 1 is what every publisher gets: a check. Steps 2 through 4 are what People Inc built on top of it. The licensing fee is the entrance ticket. The ad tech is the show.
What D/Cipher Actually Does With OpenAI
Lindsay Van Kirk, who leads D/Cipher's product development, described the technical improvement to Digiday: "LLMs are really good at helping you know that Tim and Lindsay are talking about the same thing, even though we're using different language."
Before the OpenAI integration, D/Cipher could track what people read across People Inc's 40 properties and infer purchase intent from content patterns. After integration, the system processes over 10 billion signals monthly and draws semantic connections that keyword matching misses. A reader consuming pediatric health content gets pharmaceutical ads — not retargeted ads for shoes they browsed last week.
This matters because the ad industry is moving away from third-party cookies and toward contextual targeting. People Inc isn't waiting for that transition. They're already selling it. And now, with D/Cipher+ expanding to third-party sites, they're selling it beyond their own properties — positioning the technology as an ad-targeting service that any publisher could use.
The Scorecard
| Publisher | AI Licensing (est. annual) | % of Revenue | What Else They Got |
|---|---|---|---|
| News Corp | ~$100M (OpenAI + Meta) | <1% | Cash |
| Axel Springer | ~$13M (OpenAI) | ~0.4% | Cash + ChatGPT integration |
| NYT | Undisclosed (Amazon) | — | Cash + legal leverage |
| USA Today Co. | "Notable" (Meta + others) | — | Cash |
| People Inc | ~$20–30M (OpenAI + Microsoft + Meta) | ~1.5% | OpenAI tech → D/Cipher ad platform → off-platform expansion |
Everyone got a check. People Inc got a product.
Microsoft's New Marketplace
In February 2026, Microsoft launched the Publisher Content Marketplace — a platform where publishers set licensing terms and get paid when Copilot or other AI systems ground their answers in publisher content. People Inc is one of seven launch partners, alongside Condé Nast, Hearst, Vox Media, the AP, Business Insider, and USA Today Co.
The model is usage-based: publishers get paid on "delivered value" rather than flat fees. For a top-tier brand, Microsoft projects annual royalties approaching $6 million. For a niche publication, maybe $12,000. This is the plumbing for an AI licensing market that actually scales — but at current rates, it's supplementary income, not a business model.
For People Inc, the Marketplace is another layer in their Microsoft relationship. They already have a content licensing deal for Copilot signed in November 2025. The pattern repeats: the check is modest, the relationship is strategic.
The Consolation and the Prize
AI content licensing is a consolation prize for an industry that lost something much more valuable: organic search traffic. Google sent publishers billions of free visits for two decades, then started answering the questions itself. The licensing deals are severance payments for a relationship that's ending.
The numbers make this clear. People Inc's Google traffic share dropped from 54% to roughly 24% in under three years. That lost traffic was worth far more in ad revenue than any licensing deal could replace. Even News Corp's industry-leading $100 million a year is a fraction of what Google Search used to deliver for free.
But People Inc appears to have understood this earlier than most. Instead of treating the OpenAI deal as a revenue line, they treated it as an R&D partnership. The licensing fee pays for access to technology that powers an ad product that generates multiples of the licensing revenue itself. It's arbitrage: sell your content for $20 million, use the tech you get in return to build a $50 million ad business.
Whether this works long-term depends on whether D/Cipher+ can scale off-platform — whether other publishers and ad buyers adopt a contextual targeting tool built by a competitor. That's a hard sell. But it's a better bet than hoping AI companies will keep writing bigger checks for content they can increasingly generate themselves.
THE BOTTOM LINE
The AI licensing market is real but small — a $460 million industry in 2025, headed toward the low single-digit billions by 2030. For individual publishers, the checks cover a quarter's worth of layoffs. People Inc's edge isn't the size of their deal. It's that they used the deal to build something the deal itself could never pay for.
Sources
- Digiday — Publishers cautiously count AI licensing as notable revenue (Q1 2026)
- Nieman Lab — Publishers will see no meaningful AI licensing revenue
- Digiday — How Dotdash Meredith enlisted OpenAI to boost D/Cipher
- A Media Operator — DDM sees off-platform ad targeting as big revenue driver
- Microsoft — Building toward a sustainable content economy for the agentic web
- AdExchanger — Dotdash Meredith looks to AI partnerships to protect search traffic
- AdExchanger — People Inc looks inward for growth