Publishers vs. Google

- UK publishers petitioned the CMA to curb Google's use of publisher content for fine‑tuning AI models. - They dispute Google's claim that fine‑tuning on publisher content poses 'no realistic prospect of harm.' - The move reframes AI-search conflicts as a content‑rights and platform‑legitimacy dispute (pressgazette.co.uk).

British publishers are pressing the UK’s Competition and Markets Authority to stop Google from using their content to fine-tune search AI without separate controls. (pressgazette.co.uk) The fight sits inside the regulator’s first search rules for Google under the Digital Markets, Competition and Consumers Act 2024. The Competition and Markets Authority designated Google as having strategic market status in UK search in October 2025 and opened this conduct-requirements consultation on January 28, 2026. (gov.uk) The regulator’s January proposal would let publishers opt out of Google using their material in AI Overviews and in training AI models outside Google Search, while also requiring attribution in AI results. The same package says Google Search handles more than 90% of UK general search queries and that more than 200,000 UK firms spent over £10 billion on Google search advertising in the prior year. (gov.uk) Fine-tuning is a later stage of model training: a system that already learned general patterns is fed more material so it performs better on a narrower task. Publishers say that still turns their reporting into a product input, even if the model is not simply copying whole articles back to users. (pressgazette.co.uk) That is the point of the dispute with Google’s filing. In the Competition and Markets Authority consultation documents, Google argued there is “no realistic prospect of harm” from training or fine-tuning for search AI because fine-tuning teaches processing rather than “what current information to display.” (gov.uk) Publishers told the regulator that distinction does not hold if their work improves the accuracy, freshness, and usefulness of Google’s answers. Guardian Media Group said publisher content is valuable at “every stage,” and DMG Media said bundling fine-tuning with search crawling gives Google an advantage other AI companies do not have. (pressgazette.co.uk) The publishers’ ask is narrower than a blanket ban. They want separate crawlers and separate opt-out settings for training, retrieval-augmented generation — the step where an AI system pulls in live web material at answer time — and fine-tuning, so a site can refuse one use without disappearing from ordinary search. (pressgazette.co.uk) The Competition and Markets Authority has already signaled only part of that approach. Its January proposal said it was not planning separate controls for fine-tuning, even as it proposed broader publisher choice over AI Overviews and some model training. (pressgazette.co.uk) Google has said publicly that it is working with the UK regulator and is developing more controls for sites to opt out of generative AI features in Search. The company said it wants rules that protect user experience while benefiting users, publishers, and businesses. (blog.google) What happens next is in the regulator’s hands. The consultation closed after the January 28 launch, responses were published on March 18, 2026, and the outcome will decide whether publishers can block one AI use of their work without risking their place in Google Search. (gov.uk)

Get your own daily briefing

Scout delivers personalized news, insights, and conversations tailored to your role and industry.

Download on the App Store

Shared from Scout - Be the smartest in the room.