Google Says Scaled Content Is Now a Bigger Problem Than Link Spam. Check Your AI Refreshes.
Google Says Scaled Content Is Now a Bigger Problem Than Link Spam. Check Your AI Refreshes.
At Google's Search Central Live Deep Dive in Europe, Gary Illyes said scaled content is now more of a problem than link spam. He also said Google filters out 40 billion spam pages a day, and that the spam team now uses AI to help catch more of it.
That came out the same week the September 2026 spam update was still rolling out. It started September 24 and was still active as of October 5.
If you've been running old posts through an AI tool in bulk, this one is about you.
What Google Said, and When
Three things landed within about a month of each other.
September 7: Mueller on losing faith
John Mueller answered a question about programmatic SEO by saying Google's systems may have "lost faith" in a site based on its old pages. He added that earning that trust back takes time and significant effort.
That's a site-level judgment. A pile of weak pages can drag down the good ones sitting next to them.
October 1: The generative AI guide got an update
Google updated its guide on using generative AI content and pulled in material from the Search Quality Rater Guidelines. The page now points to the rater sections on scaled content abuse and on content made with little to no effort, originality, or added value.
The core warning didn't change. Using AI to generate many pages without adding value for users may violate the scaled content abuse policy.
October 5: AI catching AI
Search Engine Roundtable reported the Search Central Live comments on October 5, citing a recap from We Are Roast. Google said AI slop is increasing and that user feedback is what new spam filters are built from. The takeaway was blunt: expect more spam updates and new spam thresholds.
Why Refreshes Can Look Like Scaled Content
Most people read "scaled content abuse" and picture a site spinning up 10,000 new city pages. Refreshes don't feel like that, because the pages already exist.
But think about what a lazy AI refresh does to 300 old posts at once. Every intro gets rewritten in the same cadence, every post gains the same "Key Takeaways" box, and every date flips to this month.
Nothing new was added. The pages just changed shape in the same way, all at the same time.
We covered why fake freshness signals stopped working back in August. A templated AI pass is the same idea, done faster and across more URLs.
How to Tell If Your Refresh Batch Is a Problem
Keep using AI if it helps. The question for each page is whether a person added something the model couldn't.
Run the "what's new" test
Pick five refreshed pages at random. For each one, write down one specific thing a reader can learn now that they couldn't learn before.
If the best you can come up with is "it reads better," that page was reworded. Put it back on the list.
Look for repeated structure
Open three refreshed posts side by side. If the intros open with the same kind of sentence and the FAQ sections ask nearly identical questions, you've made a template. Google says it now uses AI to spot exactly that kind of low-effort pattern.
Check the facts the model touched
Google's guide is direct about this. It tells publishers to manually fact-check and review all AI-generated content before publishing, including titles, meta descriptions, structured data, and alt text.
We've seen refreshed posts where the AI quietly "updated" a statistic to a number that doesn't exist anywhere. That's worse than leaving the old one in place.
What We Do Instead
Our refreshes that hold up share a few habits.
Refresh fewer pages, better
Ten pages with real updates beat 200 pages with new wording. Start with the posts losing clicks and impressions in Search Console, since those are the ones where a fix can actually show up in the data.
Add something only you have
A screenshot from your own account, or a result from a test you ran. A changed recommendation, with the reason it changed.
That's the "effort, originality, and added value" the rater guidelines describe. AI can help you write it up once you have it.
Let AI do the boring parts
Use it to find outdated references, flag broken claims, suggest internal links, and tighten a messy section. Then edit by hand, in your own voice.
Our guide on refreshing content with AI without losing your voice walks through that split.
Space out big batches
If you have 100 posts to update, don't push them live the same afternoon. Spread them out so you can watch what each group does before you commit to the next one.
FAQ
Is using AI to refresh content against Google's rules? No. Google's guidance targets using AI to generate many pages without adding value for users. An AI-assisted refresh that adds real information and gets reviewed by a person is fine.
What is scaled content abuse? It's Google's spam policy for producing large numbers of pages mainly to manipulate rankings, with little value for users. How the pages were made doesn't matter. AI, templates, scraping, and human writers can all trigger it.
Can bulk AI refreshes hurt my whole site? They can. Mueller said Google's systems can lose faith in a site based on its weaker pages, and recovering takes time and significant effort.
Is the September 2026 spam update finished? Not as of October 5, 2026. If your traffic dropped after September 24, wait for Google to confirm the rollout is complete before you judge any refreshes you shipped during it.
Know Which Pages Actually Need a Refresh
The safest refresh is the one aimed at a page that's really slipping. SEORefresher connects to Search Console and shows you which pages are losing ground, so your effort goes to the handful of posts that need it most. Find your refresh priorities at seorefresher.com.