NEWAI content checker: see which pages read like AI text

Knowledge baseTechnical seo

Black hat SEO

Every black hat SEO technique breaks one of Google's spam policies to get a page ranked: hidden text, bought links, a page for Google that visitors never see. If you look after client sites, the usual question is whether a site you take over already has some, left by a previous SEO, a plugin or a hacker.

Updated Oct 30, 20264 min read

Which techniques count as black hat SEO?

Whatever Google's list names, and more: Google "may act against any type of spam practices we detect". Since 15 May 2026 the definition also names "attempting to manipulate generative AI responses in Google Search", so AI Overviews count too. The names people use rarely match Google's:

What people call itGoogle's policyWhat it looks like
Bought links, private blog networksLink spamPaid links that pass ranking credit
CloakingCloaking, Sneaky redirectsGoogle gets one page, visitors another
Hidden text, keyword stuffingHidden text and link abuse, Keyword stuffingWhite text on white, long lists of city names
Doorway pagesDoorway abuseCity pages that "funnel users to one page"
Mass-produced pagesScaled content abuse, ScrapingA content farm
Parasite SEOSite reputation policyA third-party "best casinos" page on a medical site
Expired domainsExpired domain abuse"Casino-related content on a former elementary school site"
Hacked pages, comment spamHacked content, User-generated spamSpam others put on your site

Accordions, tabs, tooltips and screen-reader text hide things too, and they're fine. So are paid links marked rel="sponsored".

What is cloaking, and when is it allowed?

Cloaking is showing Google a different page than visitors get, "with the intent to manipulate search rankings and mislead users". The server checks the user agent (the name a bot sends along) or the IP address, and serves the crawler another page. Hackers often use it to hide the spam from the owner, who in the Japanese keyword hack even sees a 404.

The line is whether Google and visitors get the same content. Dynamic rendering is fine "as long as your dynamic rendering produces similar content", and so is a paywall Google can read, marked with paywall structured data.

Is hiding an ad-blocker notice from Googlebot cloaking? John Mueller: "Probably not. I think in general that would be fine."

How do you check a site for black hat SEO?

Search Console and one fetch test cover most of it.

I almost never see real black hat left by a previous agency. What I do see are hacked WordPress sites with spam pages the owner doesn't even see.

Start with Manual actions and Security issues, part of any Search Console audit. If either one lists something, Google penalty covers what to do next.

Look in Settings > Users and permissions too. An owner nobody at the client knows is a warning sign: in the Japanese keyword hack, "the hacker will typically add themselves as a property owner".

Then Performance: filter Query with Custom (regex) on the words spam brings (casino, pills, loans, in the site's languages), and check the Pages tab for URLs the site never made. Across four of my sites, a year of queries (up to 26 October 2026) gave three matches and 4 impressions, none of them spam. A match alone means little: open it before you worry.

In MarketingTracer's Search Insights, Explore takes the same regex on Keyword and Page, over up to 16 months, for each client site. It reads Search Console's Performance data, so it can't see what Googlebot got, the two reports or who has access.

Cloaking hides from your browser, so use URL Inspection. Click Test live URL, then View tested page, and compare the HTML and the screenshot with what you see. It's the check Google advises after a cloaking manual action.

Search Console URL Inspection live test showing the page HTML Google fetched

One of my sites in URL Inspection's live test.

For many pages, fetch each one four times: with Googlebot's user agent, as a browser, and as a browser and a phone coming from google.com. I ran it on every URL in my five sites' sitemaps: 6,536 fetches on 29 October 2026. Each page came back byte for byte the same four times, and every redirect went to the same place.

That only rules out the lazy kind. A request with Googlebot's user agent from your laptop isn't Googlebot (Google checks the real one by reverse DNS), so a hack that cloaks by IP serves it the normal page. Only URL Inspection fetches as Google.

Last, the Links report: Top linking sites and Top linking text show paid-looking domains and keyword-stuffed anchors, in a sample of the links.

Does black hat SEO still work?

Sometimes, for a while. Google finds it "both through automated systems and, as needed, human review", and caught sites "may rank lower in results or not appear in results at all". Four spam updates rolled out in 2026: 24 March, 24 June, 18 August and 24 September.

Keep reading

Free 14-day trial

MarketingTracer smart A.I. tool shows you what you need to rank. Then we'll help you get it done.

  • No credit card required
  • No installation needed
  • No strings attached
Start your 14-day trial