Judge Questions Google AI & Spam Update

This week’s SEO news has been dominated by several developments that could affect how websites approach search and AI visibility. Google is facing questions over how it uses publisher content for AI Overviews, while the latest spam update has renewed concerns about large-scale AI publishing. At the same time, Google has weighed in on the growing debate around Generative Engine Optimisation (GEO), new research has highlighted a gap between brand recommendations and website citations in AI search, and further testing has provided a more complicated picture of how ChatGPT uses Reddit.

Here are the main developments SEO professionals and website owners should know about.

Judge Questions Google’s Use of Publisher Content for AI

Google’s use of online publisher content to power its AI search features has come under further scrutiny following questions from a federal judge.

The issue was raised during a hearing in Penske Media’s antitrust case, where Judge Amit Mehta questioned Google’s description of AI Overviews as a form of “product improvement”. Digital Content Next CEO Jason Kint, who attended the hearing, said the judge appeared concerned that publishers were effectively providing the content Google uses to develop these features without having much control over how that material is used.

The judge has not yet made a decision on Google’s request to have the case dismissed.

Why It Matters

The dispute raises wider questions about the relationship between Google and online publishers. Many publishers depend heavily on Google Search for visitors, which makes choosing to block Google from accessing their content a difficult option.

At the same time, Google is increasingly using information from websites to provide answers directly within its search results. This could mean users get the information they need without visiting the original publisher.

For website owners, the case highlights an ongoing concern: publishers may be contributing valuable content to Google’s AI systems while having limited control over how that content is used or how much traffic they receive in return.

Google Spam Update Brings Large-Scale AI Content Under the Spotlight

Google’s latest spam update has also created discussion around websites publishing large volumes of AI-generated content.

Google has not confirmed that the update specifically targeted AI-written material. However, reports from some website owners suggest that heavily automated sites experienced significant visibility losses following the update.

Some SEO professionals have reported that websites using AI alongside human editing appeared to perform better, although these observations are based on individual cases rather than official evidence from Google.

There has also been discussion around Google’s S-CTS research system, which is designed to identify coordinated synthetic abuse on video platforms. However, this research relates to video platforms rather than Google Search, so there is no evidence that S-CTS was responsible for the recent search update.

Why It Matters

The bigger issue is not necessarily whether content was written by AI.

Google’s spam policies focus on content created at scale primarily to manipulate search rankings. That means websites can potentially run into problems whether the content is produced entirely by humans, entirely by AI, or through a combination of both.

The quality, purpose and method of publishing remain important factors.

SEO consultant Takuma Oka has also stressed that the examples being discussed are based on a relatively small sample and should not be treated as definitive proof of what Google changed.

For website owners, the takeaway is to avoid relying on mass-produced content simply because AI makes it possible to publish quickly. Content still needs to provide genuine value rather than being created mainly to attract search traffic.

Google Says SEO Does Not Need a Separate GEO Strategy

Google’s John Mueller has also weighed in on the ongoing debate about whether businesses need to adopt a separate approach for Generative Engine Optimisation.

When asked whether there are industries where GEO is not yet important, Mueller said that, from Google’s perspective, there is nothing particularly different that websites need to do to appear in generative AI responses within Google Search.

However, this statement needs some context. Mueller was referring specifically to Google’s own search systems. It does not necessarily mean that the same approach will work across ChatGPT, Perplexity and other AI search platforms.

Why It Matters

For businesses already investing in traditional SEO, Mueller’s comments suggest there may not be a completely separate technical checklist required for Google’s AI-powered search features.

The fundamentals of good SEO still matter, including useful content, clear information and a technically sound website.

However, different AI platforms can use different retrieval and ranking systems. As a result, a website may perform differently across Google’s AI features and other generative search tools.

One way of looking at it is that the core SEO principles remain similar, while the way AI systems retrieve and select information can vary.

AI Search Can Recommend a Brand Without Citing Its Website

Another interesting development comes from new ecommerce research looking at how AI search tools recommend brands and choose the sources they cite.

Shero Commerce analysed 1,851 citations across Google AI Mode, ChatGPT and Perplexity. The research found that only 2.8% of citations came from brand-owned websites.

More interestingly, when an AI system recommended a particular brand by name, the brand’s own website was used as the citation only 31% of the time.

In many cases, another website was used as the source instead.

Why It Matters

This highlights an important difference between being mentioned by an AI system and receiving the actual website citation.

A company could therefore gain visibility because an AI tool recommends its brand, while the user is directed to a third-party website such as a marketplace, publication or review site.

That creates a challenge for businesses measuring their AI search performance.

Simply tracking brand mentions may not provide the full picture. Website owners also need to understand where citations are coming from and which websites are receiving the clicks generated by AI search.

The research shows where citations appeared, rather than explaining exactly why an AI system selected one website instead of another. Therefore, it should not be treated as proof that changing a brand’s own content would automatically increase citations.

ChatGPT’s Use of Reddit May Depend on the Search Query

There has also been further testing into how ChatGPT uses Reddit when retrieving information.

Earlier data suggested that Reddit’s share of ChatGPT Search citations had dropped significantly during August. However, new testing indicates that Reddit is still being used during the retrieval process and can still account for a large proportion of the final citations for certain searches.

In one test, ChatGPT specifically searched the subreddit r/whatnotapp. Out of 71 pages retrieved by the system, 48 came from Reddit threads. Six of the eight sources ultimately cited in the response were from that subreddit.

Another test carried out several days earlier produced a very different result. ChatGPT retrieved 84 Reddit threads for a query but did not cite any of them in the final response.

Why It Matters

The difference between these tests shows why citation data alone may not tell the complete story.

An AI search system can retrieve and examine a website without eventually using it as a visible source in its answer.

It also suggests that ChatGPT should not necessarily be described as either favouring or avoiding Reddit based on a single set of searches. The behaviour appears to depend heavily on the individual query and the information being requested.

For businesses and SEO professionals, this is another reminder that AI search behaviour can be much less predictable than traditional search results.

The Bigger Picture: Being Visible Does Not Always Mean Getting the Click

Taken together, this week’s developments point towards a growing separation between visibility and value in AI search.

A publisher can provide information that helps an AI system answer a question without receiving a website visit. A brand can be recommended by an AI tool while a different website receives the citation. Reddit can be heavily used during the retrieval process but may not always appear among the final sources.

The GEO discussion follows a similar pattern. Being discovered by an AI system is only one stage of the process. For businesses, the more important questions are whether they are mentioned, whether their website is cited, whether users click through and, ultimately, whether that visibility leads to enquiries, sales or other useful outcomes.

As AI search continues to develop, simply measuring how often a website is crawled or how frequently a brand appears in an AI answer may no longer be enough.

SEO reporting is likely to become more focused on the full journey — from being discovered and retrieved to being mentioned, cited, clicked and converted.