Recommended Articles
Tech & AI

Reach Strikes Google AI Content Deal — and Sends OpenAI a Cease-and-Desist

By Angeled · October 16, 2026 · Tech & AI
Reach Strikes Google AI Content Deal — and Sends OpenAI a Cease-and-Desist
Reach Strikes Google AI Content Deal — and Sends OpenAI a Cease-and-Desist

Reach plc, the UK publishing group behind the Daily Express, Daily Star, and OK! Magazine, has agreed a deal with Google giving the tech giant paid access to expanded rights to use news content from Reach's UK and Ireland portfolio for its news AI pilot. In the same move, the publisher revealed it has sent OpenAI a cease-and-desist letter — sharpening the industry's two-track approach to AI companies: pay us, or face legal pressure.

The Google deal

The agreement, understood to run for two years, gives Google expanded rights to Reach's news content for use in experimental AI features and new ways of delivering news. Google says more than 200 publications globally have joined its news AI pilot program, alongside its Google News Showcase program and a pay-per-value AI contribution pilot for smaller publishers. For Reach, CEO Piers North is positioning the agreement as a model for future LLM partnerships.

The OpenAI cease-and-desist

Reach's posture toward OpenAI is the opposite of its posture toward Google. The publisher says it has sent OpenAI a formal cease-and-desist letter, stepping up pressure on AI firms it believes are using its content without a deal.

The scale of the underlying complaint is striking: Reach says it has logged more than 100 million successful scrapes of its content across its portfolio over the last six months alone. That figure captures automated harvesting of Reach's journalism — presumably by AI crawlers and other bots — at industrial scale.

The publisher-vs-AI-lab landscape

Reach's dual approach — license to one lab, litigate-pressure another — sits inside a fast-moving global standoff between publishers and AI companies over training data.

The general shape of the conflict: AI labs need vast amounts of text to train and operate their models, and the open web — including copyrighted journalism — has been the cheapest source. Publishers argue this uses their work without permission or payment; AI companies counter with fair-use/fair-dealing arguments and the claim that training is transformative use of public material.

The New York Times' litigation against OpenAI and Microsoft is the highest-profile test of these claims and is being watched as a potential precedent-setter, though its outcome and timeline remain uncertain. Alongside the courtroom track, a parallel market track has emerged: paid licensing deals between publishers and AI labs. What makes Reach's move notable is the explicit linkage of the two tracks: announce the Google deal as the "model" partnership and the OpenAI cease-and-desist as the alternative — a negotiating strategy played out in public.

What "pay-per-value" models look like

One of the most interesting details is Google's pay-per-value AI contribution pilot for smaller publishers. Traditional licensing has been built on flat fees or per-article rates. Pay-per-value attempts something different: compensating publishers based on the measured value their content contributes to AI outputs — in principle, payments tied to how often content is retrieved, cited, or used in generated answers. The methodology is still being defined, but the direction matters: the industry is groping toward a metered, usage-based royalty model for the AI era — closer to music streaming royalties than to old syndication deals.

How content businesses should think about AI licensing

Reach's playbook offers a practical template for any content business — from national publishers to niche B2B media and even large blogs — navigating the AI era:

  1. Audit your exposure. Reach's 100M-scrape figure came from measurement. Instrument your sites to quantify crawler activity before negotiating or threatening — data is leverage.
  1. Segment your counterparties. Reach's two-track approach — deal with the willing, pressure the unwilling — monetizes cooperation while preserving legal optionality. A blanket "no AI" stance leaves money on the table; a blanket "yes" surrenders leverage.
  1. Define the rights precisely. Two years, defined use cases, experimental scope: time-boxed, use-case-specific licenses preserve your ability to renegotiate as the market matures.
  1. Watch the small-publisher models. If metered royalties become standard, even niche content businesses could earn recurring AI-licensing revenue — provided they have the rights and measurement in place.
  1. Model the legal track honestly. Cease-and-desist letters and litigation are expensive and slow; they make sense as leverage with scale and clear evidence. Smaller publishers should weigh coalition approaches — industry associations, collective licensing bodies — over solo action.

Why it matters for founders and investors

The AI-content licensing market is becoming a real revenue line, not a legal footnote. For investors, publishers with large, rights-cleared archives are suddenly holding appreciating assets — every LLM lab needs training and retrieval content, and the supply of willing licensors is finite. Media startups that build clean rights chains from day one (clear contributor agreements, crawl policies, licensing infrastructure) are building balance-sheet value that acquirers and partners will pay for.

For founders in AI itself, the lesson runs the other way: unlicensed scraping is becoming a contingent liability. The startups that will scale through this era are the ones that treat content licensing as a cost of goods — budgeted, contracted, and disclosed — rather than a legal risk to be discovered during due diligence. Expect investors to ask about your data provenance with the same rigor they once reserved for cap tables.

Angeled.com connects founders with investors.

Sources: Digiday (Oct 16, 2026)

Read Next