Google spam update enters second phase as AI content review becomes "critical"

Phase two of the September 2026 spam update lands mid-week

" Google's own framing of the September 2026 Spam Update is that it is improving how it detects search spam, and an industry practitioner thread describes the change as a global algorithmic update applying to all languages and regions, aimed at neutralising content that violates Google's policies. Early analysis in an Instagram reel clip points to pages that are "heavily templated, programmatic, or AI-generated" as the surface area being penalised, which is the same signal other SEO commentators are watching. For practitioners, the practical effect is movement in organic rankings and traffic during a still-active rollout, not a one-day event to react to and forget.

Google now calls AI fact-checking "critical"

On October 2, 2026 Google updated its Search guidelines with new language on how to ensure accuracy, authenticity and trustworthiness prior to publishing generative AI content. The Search Engine Roundtable recap reports that Google now says it is "critical to manually review all AI-generated" material before it ships, which is a notable hardening from earlier guidance that framed such review as recommended. A separate industry post frames this as guidance for "anyone publishing" generative AI material, not just large publishers.

The operational takeaway is that AI-assisted publishing workflows need a documented human review checkpoint; "we used an LLM and edited the output" is no longer enough to describe that checkpoint.

AI Overviews reappear for brand-name queries

Google is returning AI Overviews for a new policy on misuse of support channels, and AI Overviews have returned for many large brand names, per the Search Engine Roundtable video recap of October 2, 2026. Returning to brand queries matters because brand-name SERPs are the queries a site's own audience runs; an AI Overview above the fold on those queries changes what users see before they ever reach the brand's own listing. Combined with the spam update, this means two separate surfaces are in motion at once: a punitive surface for thin and AI-generated pages, and a generative surface that is re-expanding over navigational queries.

"Scale via AI" is the warning, not the playbook

A practitioner post from October 2, 2026 by an SEO and AI Search consultant frames the spam update as evidence that scaling content via AI remains risky, writing "for those still wanting to scale via AI" with 578 engagements at the time of capture. The post is not a publisher's announcement, but it captures the sentiment circulating across the SEO community in the wake of phase two: the update is treating heavy templated and AI content as the primary target. For teams running AI-assisted content programmes, the immediate question is no longer whether Google can detect such content but whether the new guidelines require a review step they do not yet have in their pipeline.

What to check before the next reporting cycle

The spam update is still rolling out, and Google's own communications describe the change as ongoing rather than complete, so ranking and traffic volatility is expected to continue rather than resolve on a specific day. Concrete checks a reader can run over the coming week: pull the dates of the most volatile pages in Search Console and separate movements on Wednesday and Thursday from earlier-week movement; flag any templated or programmatic pages that lost visibility and decide whether they meet the new manual-review bar; audit AI-assisted drafts for the fact-checking checkpoint Google now describes as critical; and track whether brand-name queries in your category are returning AI Overviews, since that surface is back and changing above-the-fold behaviour on your own audience's queries.

GEO Brand Question Generator is useful here for stress-testing whether brand queries in your category still surface your own assets inside AI answers.