What's your site's AI Spam Score?
A 0–100 score of how strongly your website matches the scaled-AI-content footprint that Google's June 2026 spam update targets — computed from 40+ public, explainable signals across your whole site.
Free · No signup · Full 7-dimension breakdown in ~1 minute
One number for a site-wide pattern
Google doesn't demote content for being AI-written — it demotes scaled content abuse: many pages produced primarily to rank, not to help anyone. That's a production pattern, and it's visible from the outside: templated pages that share a skeleton, publishing bursts in your sitemap, thin clusters with no authors, no sources and no reason to exist.
The AI Spam Score measures exactly that pattern. PureRank crawls a sample of your site, reads your sitemap history, and scores seven dimensions — AI language markers, stylometry, formatting patterns, originality & templating, publishing velocity, the trust/E-E-A-T surface, and AI-search manipulation. Every signal is public and explained in the report; the full methodology is open.
Per-paragraph detectors guess authorship — independent tests put real-world accuracy around 72–74%, with false positives that hurt honest writers. Authorship also isn't the question Google asks. The question that predicts risk is whether your site looks mass-produced — and that needs site-level forensics, not a text box.
Seven dimensions, weighted and explained
Each dimension is a risk score of its own — the report shows every signal that fired and why.
Language & style
250+ LLM stock phrases, giveaway artifacts, sentence-rhythm uniformity and vocabulary diversity — the linguistic residue of unedited model output.
Templating & scale
Cross-page similarity, reused heading skeletons, doorway-style slug clusters, and sitemap forensics: publishing bursts and index bloat are the signature of "scaled".
Trust surface
Authors, dates, sources, contactability, schema.org, domain history — the context that decides whether AI-assisted content survives a quality re-rating.
Six site profiles (publisher, e-commerce, docs, local business, personal blog, landing) re-weight the dimensions so you're judged against the right baseline.
What the score is not
- Not a Google metric. PureRank is independent; the score estimates risk against Google's published spam policies, not their internal systems.
- Not proof of authorship. No tool can prove who or what wrote a text. We calibrate hard against false positives and show our work per signal.
- Not a verdict on AI use. AI-assisted content that genuinely helps people is fine — and scores fine. Mass-produced filler is what lights the score up.
If your score is high, the report tells you which dimension drives it and lists the exact pages — so the fix is targeted, not a guess. Recovering after an update? Start with the spam-update recovery guide.
Straight answers
Does a high AI Spam Score mean Google will penalize my site?
Not automatically. The score estimates how strongly your site matches the scaled-content-abuse pattern Google's spam systems target. A high score means elevated risk — and tells you which dimension drives it (templating, publishing bursts, thin clusters, weak trust surface) so you can fix it before an update decides for you.
How is this different from an AI detector like GPTZero or Originality.ai?
Per-text detectors answer "was this paragraph machine-written?" — unreliably, with real false-positive rates. PureRank answers a different question at the site level: "does this site look mass-produced rather than made for people?" That production pattern — not authorship — is what Google's spam policies actually act on.
Is the check free?
Yes — a scan takes about a minute, needs no signup, and shows the full 7-dimension breakdown. Paid plans add deeper crawls, private reports, daily monitoring and API access.
Can the score prove content is AI-written?
No, and we say so plainly. It's a heuristic estimate built from 40+ public, explainable signals — use it to prioritize a human review, and read the per-signal breakdown rather than the single number.
Check your AI Spam Score now
40+ signals, 7 dimensions, about a minute. Then embed your score as a live badge.
Run a free scan →