- Window
- ~2026-09-24 → 2026-09-25 during early spam-update rollout.
- Scope
- Classic SEO and GEO: Google AI Overviews and AI Mode, Bing Copilot and AI Search, ChatGPT Search, and Perplexity. No Chinese or Baidu sources.
- Status
- September 2026 spam update in progress (started 2026-09-24 09:15 US/Pacific; Google said up to two weeks).
- Confidence
- Publisher
Summary
- SEO: Spam update early days; expect multi-day GSC reads.
- GEO: SEJ clarifies AI Overview link types vs Search Console counting.
- Research: SAFE paper on synthetic abuse — interesting, not confirmed as this spam update’s named mechanism.
01
September spam update: early volatility window (Unconfirmed amplitude)
Publisher
Rollout itself is Official; which sites/how hard hit is industry reporting.
What happened
With the spam update live since Sep 24, Roundtable/SEL later mark Sep 25–27 as an early volatility wave. Google still has not named targeted spam tactics beyond a normal spam-systems improvement.
Why it matters
Use GSC multi-day patterns, not single-day rank-tracker spikes (trackers later reported noisy SERP returns during this rollout).
Source
- Google Search Status — September 2026 spam update incident
- Search Engine Roundtable — Google September 2026 Spam Update Weekend Impact
- Search Engine Land — Google September 2026 spam update done rolling out Retrospective volatility windows including Sep 25–27.
02
SEJ: not every AI Overview “link” reaches the open web
Publisher
What happened
SEJ (2026-09-25) explains Google’s link displays in AI Overviews/AI Mode (inline links, end suggestions, discussion previews, follow-ups into AI Mode) and GSC rules: external-page clicks count; query-refinement links do not count as clicks/impressions for that link. Relays Peec’s ~26% in-text external-link chart and AI-Mode-opening anchors.
Why it matters
Core GEO measurement literacy: generative AI impression reports ≠ click reports; underlined UI ≠ always an outbound citation.
Source
03
Google research: SAFE multi-agent detector for synthetic “AI slop”
Publisher
Paper says deployed; SEJ speculation that it may relate to the spam update is not Google-confirmed.
What happened
SEJ (Roger Montti, 2026-09-25) covers Google’s short paper on SAFE (Scaled Abuse Forensics Examiner): multi-agent forensics for coordinated synthetic media / “AI slop,” including behavior, content, and channel-cluster agents. Paper claims early deployment sped investigations; no public metrics and no Status Dashboard link tying SAFE to the September spam update.
Why it matters
Useful context for spam / AI-generated abuse defense—but do not equate the paper with the confirmed spam update without Google saying so.
Source
- Search Engine Journal — Google Has Deployed A New AI Spam Detector Called SAFE
- Google Research — The Synthetic Gap … SAFE Paper landing; exact public date may match SEJ coverage window.
04
Status Dashboard: spam update still rolling
Official
What happened
Incident remains open/in progress; no completion notice yet.
Why it matters
Official state for client reporting this week.
Source
Also noted
| Signal | Date | Note |
|---|---|---|
| AI features site-owner doc | checked | No material change |
Method notes
- P0 sites polled via WebFetch/WebSearch; Status Dashboard checked for confirmed ranking/spam state.
- Rumors kept Unconfirmed until Google/Bing/official confirmation.
- No Chinese/Baidu sources.