Key takeaway: Google faces UK AI search opt-out rules while US publisher suits were dismissed and Ahrefs reports disrupted SEO data.
Key Takeaway
Google’s AI search position is developing along three separate tracks. In the United Kingdom, the Competition and Markets Authority required Google to let publishers opt out of having their content power AI features in Google Search and to opt out of its use for AI-model fine-tuning. In the United States, a federal judge dismissed two publisher lawsuits over AI Overviews on September 30, leaving those plaintiffs without the court-ordered choice sought in the cases. Separately, Ahrefs said on October 22 that Google had been showing swapped search results to suspected SERP data providers, with the affected share ranging from 5% to 20% depending on the provider. Together, the developments affect publisher content controls and the way marketers verify SEO search visibility.
What UK publishers can control in Google AI search
The CMA’s binding conduct requirement, imposed on June 3, gives publishers a route to reject the use of their content for Google’s search AI features and for AI-model fine-tuning. The underlying controls serve different purposes. Google-Extended governs whether site content helps train Google’s AI models, and Google’s documentation says it does not affect Google Search. A separate setting for Search generative AI features can remove a site from AI Overviews, AI Mode and AI features in Discover, but it does not control AI training.
Google has nine months from June 3 to implement the CMA changes and must submit compliance reports every six months during the first year. The requirement therefore concerns a specific regulatory remedy in Britain, not a global confirmation that every publisher can already block every use of its content. It also leaves publishers to weigh potential AI search exposure against content-use preferences and search traffic.
Why the US publisher cases did not create an AI opt-out remedy
US District Judge Amit Mehta granted Google’s motions to dismiss the Chegg and Penske Media Corporation cases in full. Penske owns Rolling Stone, Variety and Billboard and sued with 14 affiliated publishers. The judge found that the publishers had alleged an expectation of traffic rather than an agreement with Google, and that they had not plausibly shown AI Overviews to be a separate product from Google’s search engine for antitrust purposes. The dismissal was without prejudice and can be appealed. A similar case from the Helena World Chronicle had already been dismissed in March.
The opinion acknowledged publisher harm but left the broader policy questions to Congress. That means the US ruling did not impose the kind of publisher choice required by the UK regulator. It also does not resolve how publishers should manage content used in Google’s AI search products.
Why SEO ranking reports now need additional verification
Ahrefs said that, over the previous few weeks, Google had shown different results to suspected SERP data providers. It described the issue as industry-wide and put the affected share between 5% and 20%, depending on the data provider. Ahrefs said it was filtering and rejecting abnormal pages, requesting accurate replacements, and cleaning up older results that might have been affected.
The company did not publish a sample size, a start date, a provider-by-provider breakdown, filter methodology, or the number of historical records changed. It also did not identify which Ahrefs products use affected data or say whether corrected points will be marked in the interface. For agencies, brands and publishers, a reported SEO drop during September or October may therefore require a second check rather than an immediate conclusion about actual search visibility. Google’s September spam update overlapped with the period in which vendors reported swapped pages, adding another reason to compare third-party rank data with Search Console and other independent measurements.
Method and Limitations
Readers can verify the UK development against the CMA requirement and its implementation timetable, review the US opinion for the dismissal grounds, and compare Ahrefs reports with first-party Search Console data and independent ranking sources. The available accounts do not establish that every query, country, device or website was affected. They also do not show how widely UK publishers will use the opt-out settings. Google did not publicly respond in these reports to vendors’ accounts of swapped results. The US cases may still be appealed, and Ahrefs has not specified the reach of its historical cleanup.