Google caught feeding fake SERPs to rank trackers, peaking at 34% of results
daluoseo · x · 2026-10-19
Nozzle's analysis of 27.5 million SERPs across 49 countries shows Google has been serving poisoned results to third-party scraping tools since mid-September: requests failed en masse from Sep 13, poisoned pages appeared from Sep 20, and by Sep 27–29 about a quarter of results were fake (peaking at 34% on Sep 30 per Keyword.com data), still running at 5–11% daily in October.
The fake pages look completely normal—same templates, no errors or captchas—except real top-ranking sites drop to 20% visibility and slots are filled with Reddit, Facebook, LinkedIn and Quora results, which tools ingest as genuine data.
Combined with Google killing num=100, rolling out goto links, suing SerpApi, and Reddit's request to block scraping, the real target appears to be companies feeding Google results to AI systems. SE Ranking says it fixed historical data (2% still affected); Semrush says it added validation with limited impact beyond page one.
Related event: Google Caught Serving Fake SERPs to Rank-Tracking Tools(3 posts)→
More from AGI Musings
- Gensyn co-founder to discuss auditable training, AI free of few-company control — benfielding · 2026-10-20
- He pitched pretraining-inference compute exchange rate research 9 months before o1 — RylanSchaeffer · 2026-10-20
- Author of 'Emergent Abilities a Mirage' says results valid but obvious to labs — RylanSchaeffer · 2026-10-20
- AI fluency boils down to reframing business bottlenecks as prediction problems — sanderssays · 2026-10-20
- Stewart Squared Ep. 111: When Robots Become Agents — StewartalsopIII · 2026-10-20
- AI rewards laziness over hard work — ask better questions, start from 80% — EXM7777 · 2026-10-20