This week is about inputs and attribution moving faster than most dashboards. Anti-spam systems are tightening, Search Console is breaking out image-driven and AI-assisted traffic, developer docs close a loophole on CSS-generated text, and ChatGPT’s bots want a clear path. Momentum is up and to the right for multimodal queries; audits that used to be quarterly now need to be routine. Start with the shift that can yank visibility overnight, then make space in reporting for the modalities that were invisible yesterday.

Google’s September 2026 spam update is live worldwide

Google’s Search Status Dashboard lists a “September 2026 spam update,” released Sept 24 at 09:15 PDT. The incident affects Ranking and applies globally in all languages. Google said this is a spam-systems improvement. During rollout, sites that violate spam policies can see reduced visibility. Expect movement while the update propagates.

Action list: audit templates and high-risk sections for spam policy issues, tighten moderation, and track ranking/traffic shifts by spam-sensitive segments.

What it changes: Watch for volatility and audit against Google’s spam policies; remediate issues and monitor while the rollout continues.

Search Console now attributes Lens and AI-assisted traffic

Google introduced a new “web multimodal” search type in Search Console performance reports. Traffic is now attributed from Lens, Circle to Search, image uploads, and Chrome’s Search this image, and visibility from generative AI features is included. Rollout began Sept 24 globally. Data appears once your site actually receives queries from these entry points, per the post by product managers.

Set up saved filters, compare against Web and Image types, and break out appearance metrics tied to generative features. This gives a cleaner view of non-typed search demand.

What it changes: Add the Web multimodal type to reporting, segment by source, and adjust content tracking for Lens and Circle to Search queries.

Text injected via CSS content may not be indexed

Google updated its developer documentation to clarify that text placed via the CSS content property may not be indexed, because that text isn’t part of the DOM. If you’ve been relying on::before/::after to inject headings, captions, or disclaimers, search may not see them.

Put meaningful, indexable copy in the HTML, not CSS-generated pseudo-elements. Use CSS content for decoration only. This is a clarification to the docs, not a new feature, but it has real consequences for sites that styled away essential text.

What it changes: Review templates and move any required text from CSS pseudo-elements into the HTML.

OpenAI asks sites to allow AdsBot and recommends SearchBot access

OpenAI updated its advertiser guidance on allowing web crawlers. The help article instructs sites to allow OAI-AdsBot so ChatGPT ads can validate landing pages, and recommends allowing OAI-SearchBot for appearance in ChatGPT search results. It provides robots.txt examples, notes on CDN/WAF allowlisting, and steps to re-submit if crawls were previously blocked.

If you block these bots, ChatGPT ads validation and search presence may be affected. That guidance includes how to allowlist via common CDNs and WAFs.

What it changes: If you advertise in ChatGPT or want inclusion in ChatGPT search, update robots.txt and allowlist the bots, then re-submit as directed.