Our Take
This is a regulatory win for publishers, not a technical one: Google already has the infrastructure to exclude content; the CMA simply made it mandatory and visible.
Why it matters
Publishers have been losing negotiating leverage as Google scaled AI Search without consent mechanisms. The UK ruling sets a legal template that other regulators (EU, US) will likely follow, forcing AI companies to embed publisher controls as standard rather than optional.
Do this week
Publishers: audit your Search Console access this week and document which content feeds AI Overviews today, so you can make informed opt-out decisions before Google rolls out the toggle globally.
UK regulator mandates publisher controls on Google AI Search
The Competition and Markets Authority (CMA) imposed new conduct requirements on Google Search requiring the company to give publishers direct control over whether their content appears in AI features like AI Overviews and AI Mode. Publishers can also prevent Google from using their content to fine-tune AI models.
Google must roll out a toggle in Search Console that lets website owners manage content inclusion in generative AI search features. Websites that opt out entirely will receive no traffic or impressions from AI Search features. The company also committed to properly attributing publisher content in AI-generated results using clear links.
Google is already testing these features with a subset of UK website owners and plans to deploy them globally after the pilot phase. New Search Console insights will show publishers which webpages appear in AI responses and in which countries.
The ruling came after Google had reportedly rejected earlier proposals to give publishers more granular control, citing the space as an evolving monetization opportunity. The CMA framed the requirement as leveling negotiating power between Google and news organizations ahead of content licensing deals.
This establishes a regulatory precedent for AI content attribution
The CMA ruling creates enforceable law around publisher consent for AI training and inference, not merely guidance or best practice. That distinction matters because it removes Google's discretion to deprioritize the feature or bury the control. Publishers now have a legal guarantee of access to the mechanism.
Equally important: the ruling does not require Google to pay for AI Search inclusion. It only mandates visibility and control. That sidesteps the harder question of whether publishers should be compensated. Instead, it shifts the negotiating table: publishers can now credibly threaten to opt out, giving them leverage in licensing deals with Google and other AI vendors.
Regulators in the EU and US are watching. The Digital Markets Act already subjects Google to stricter compliance regimes in Europe. US lawmakers have signaled concern about AI training on copyrighted content. This UK ruling provides a working example of how to operationalize publisher consent at scale without breaking the underlying product.
What publishers and platforms should do now
Publishers should begin cataloging which content currently feeds into AI Search features. Google's new Search Console insights will help, but having your own baseline matters for assessing the traffic impact of opting out.
For AI vendors building RAG or retrieval systems: expect similar requirements to spread. If you are ingesting publisher content, plan for opt-out mechanisms now. The CMA's logic—that creators own decisions about derivative uses—will likely extend to fine-tuning, training data filters, and other content pipelines.
For platforms considering AI features over user-generated or licensed content, the cost of retrofitting consent and attribution controls is now visible. Baking them in at design time is cheaper and faster than defending their absence to regulators later.