Google September Spam Update Enters Final Phase, AI Overviews Expand to Brand Name Searches

Good morning. This is the SEO news for October 10, 2026. This morning, we will cover the rollout status of the September spam update, tracking data on the appearance rate of AI Overviews in brand name searches, a subtly useful improvement to Search Console, and two document updates released by Google.

1. Google September Spam Update, Third Phase (10/4-6) Appears to be the Final Phase

The “September 2026 spam update,” which began on September 24, 2026, differed from previous updates that typically finished in about two days, as Google explained from the start that it would “take about two weeks to roll out.” The actual fluctuations were not continuous but were observed in multiple waves.

The reported timing of the fluctuations includes the first wave from September 25-27, the second wave on September 30, and the third wave from October 4-6. This third wave is considered to be the final phase. Note that as of the time of writing, no official announcement of completion has been confirmed from Google, so it is safe to treat this information as being at the “appears to be complete” stage.

Additionally, there are observations that fluctuations are continuing even after the update, and there is talk of an increase in reports of manual actions. However, it has not been confirmed whether there is a causal relationship between the update and the increase in manual actions. At this point, this should be taken as a hypothesis.

– Source: Search Engine Roundtable
– Source: Search Engine Roundtable (Weekly Summary)

Practical Note: Because the fluctuations occurred in multiple waves, if you only look at “around September 25,” you might miss the impact. It is safe to check the entire period from September 24 to October 6 as a single timeline. The Search Console “Country” filter improvement mentioned later can also be used for this period comparison.

2. AI Overviews, Brand Name Search

Appearance Rate Significantly Increases (Data from Multiple Trackers)

Search Engine Journal has summarized three surveys tracking the appearance rate of AI Overviews in brand name searches in September 2026. The premise that “AI Overviews are unlikely to appear in branded searches” has changed, at least in terms of measurement.

Ahrefs tracked 100 brand names selected from Interbrand’s “Best Global Brands 2025” every week from July to September. They reported that while only 5 out of 98 valid cases (about 5%) had AI Overviews as of July 1, they confirmed AI Overviews in 63 out of 100 queries by September 30. Of those 63 cases, 48 were displayed in what the company calls “Position 1.” The most frequently cited domains were Wikipedia (71.9%), YouTube (38.6%), and LinkedIn (22.8%).

In another survey, Chris Long (Nectiv) retrieved 100 major brands using SerpApi on September 30 and confirmed AI Overviews in 93 cases. DemandSphere’s daily measurements showed an increase from 26.12% on September 1 to 90.48% on September 27, with 82.06% reported on September 29. A broader crawl of US desktop results by Ahrefs also indicated that AI Overviews appeared for 82.91% of brand-related keywords as of September 29.

However, there are some points to note. These are all appearance rate data; this data does not reveal the actual impact on click-through rates, the extent to which the brand’s own site was cited, or whether this is the date Google officially announced the rollout.

– Source: Search Engine Journal

Practical Note: You can track future changes by your own standards by checking the results of searches for your own brand name, recording whether AI Overviews appear, and if so, which sites are cited. The trend that Wikipedia, YouTube, and LinkedIn are the top citation sources provides material for considering where to maintain brand information.

3. Search Console’s “Country” filter allows multiple selections

Support for multiple selections

You can now select multiple countries simultaneously in the “Country” filter of the Search Console performance report. Previously, you could only select one country at a time, so if you wanted to view multiple countries together, you had to export the data to CSV and aggregate it yourself.

With multiple selection enabled, you can now view metrics on-screen for a combined region, such as the German-speaking area (Germany, Austria, and Switzerland). A key advantage is being able to see metrics calculated by Google that cannot be accurately derived through simple addition, such as average position.

– Source: Overseas SEO Information Blog (Kenichi Suzuki)

Practical Note: If you operate a multi-language or multi-region site, it is practical to create segments for “major countries” for regular monitoring. While this may have limited use for sites focused solely on the Japanese market, it is helpful when you want to observe the impact of algorithm updates by regional groups.

④ Google publishes documentation for the “UGC Fresh Data Program”

Google has published documentation for the “UGC Fresh Data Program” on Search Central for platforms that handle user-generated content (UGC). It is described as a mechanism where approved platforms can send content and interaction signals to Google through a dedicated data ingestion pipeline, distinct from the standard Indexing API.

Participation requirements include: UGC or forum posts must be published on dedicated pages with stable URLs; there must be a sufficient volume of content and user scale; content must be publicly accessible without login or paywalls; implementation of OAuth 2.0 authentication and JSON-LD; active moderation and reporting mechanisms; and content must be submitted as quickly as possible (ideally within minutes). Updates to engagement counts are also required to be sent within 72 hours of content creation.

Applications are submitted via a form, with an estimated response time of 6 to 8 weeks. Google explicitly states that “this program does not guarantee that content will appear in search results.”

– Source: Google Search Central
– Source: Search Engine Roundtable

Practical Note: This is intended for UGC platforms like Reddit and is not a portal for general corporate sites or media to speed up indexing. It should only be considered if your company has community features, Q&A sections, or a large volume of reviews.

⑤ Google adds descriptions regarding “human fact-checking” and fake author information for AI-generated content

On October 1, Google added descriptions regarding accuracy to its Search Central guide on the use of generative AI content. After explaining that generative models do not retrieve facts but rather predict the next likely sequence of words from training data—and thus may contain errors (hallucinations)—it emphasizes that human fact-checking and review of AI-generated content before publication is essential.

This applies to more than just the main body text. It is explicitly stated that metadata that may appear in search results, such as

elements, meta descriptions, structured data, and image alt attributes, are also within the scope of review. It also touches on disclosing the use of automation in a way that is appropriate for readers, noting that informing readers about how content was created provides necessary context.

In related news, it was reported that on October 6, a statement was added to the helpful content documentation advising against the use of fake author information. The content states that fabricating creator profiles to make it appear as though they were written by human experts—using AI-generated headshots, fictitious names, and false credentials—is a form of deception.

It should be noted that Google has not stated that the ranking mechanism has changed. It is reasonable to interpret this update as a formalization of existing concepts into documentation.

– Source: Google Search Central
– Source: Search Engine Roundtable

Practical Note: In environments where AI is used to mass-produce articles, review processes tend to focus heavily on the body text. Please ensure that titles, descriptions, structured data, and alt attributes are also included in the review process. Regarding author profiles, the straightforward approach is to accurately list the names, titles, and backgrounds of real staff members rather than creating fictional personas.

Summary for Today

– The September spam update is considered to have concluded with its third phase from October 4 to 6. Please check for impact across the entire period from September 24 to October 6.
– The appearance rate of AI Overviews in brand name searches has increased significantly according to multiple measurements. However, the impact on clicks remains unknown.
– Google’s AI-generated content guide now explicitly requires human fact-checking before publication, including metadata.

Have a great day.