What Are White-Hat, Gray-Hat, and Black-Hat GEO Tactics?

How do the three categories break down?

Why the line falls where it does is covered in whether AI visibility optimization is ethical.

Can ranking parasite pages on Google improve ChatGPT and Claude visibility?

Sometimes briefly, and at growing risk. "Parasite" pages are content placed on a high-authority site that is not yours, such as a sponsored section of a news site or a coupon directory hosted by a publisher. The aim is to borrow that site's ranking strength. Because many AI answers draw on search results, a parasite page that ranks can be retrieved and cited. ChatGPT's search draws partly on third-party search providers, and Claude's on Brave's index.

The problems:

• Google treats it as spam. Its site reputation abuse policy covers third-party pages published mainly to exploit the host site's ranking signals. In November 2024 Google clarified that this applies regardless of any first-party involvement or oversight. Host sites have been hit with manual actions, and many removed such sections.
• The benefit is fragile. When the host removes the section, the citations disappear; when the search engine demotes it, they become much less likely.
• It builds no lasting asset. The page belongs to someone else, and assistants attribute the content to the host, not to you.
• Readers notice. A recommendation traced to paid filler on a news site undermines the recommendation.

Genuine placements are different: an expert article under your name in a relevant publication, commissioned by its editors, or disclosed sponsored content that meets the publisher's standards. They earn visibility on their merits rather than by exploiting a host's reputation.

What is context stuffing in AI search, and why can it hurt a business?

Context stuffing means packing pages with material meant to maximise what gets retrieved into an AI model's context, rather than to help readers. Common forms:

• Long lists of keywords, cities or product variants added to pages.
• Dozens of near-duplicate questions and answers, each phrased slightly differently to match more prompts.
• Repeated pairings of the brand with category terms, such as "best CRM" in every paragraph.
• Hidden text or instructions visible only to machines.

It hurts in several ways:

• Search engines penalise it. Google's spam policies list keyword stuffing, including blocks of text listing cities a page wants to rank for, and hidden text. A demoted page is much less likely to be retrieved, and a page removed from the index is not retrieved at all.
• It does not work well. The research that coined "generative engine optimization" found keyword stuffing performed poorly compared with adding sources, quotations and statistics.
• It dilutes the passages that matter. Retrieval works on passages. Filling a page with variants makes each passage less specific and less quotable.
• It degrades long contexts. Research on long inputs has found models use information at the start or end of a long context better than information buried in the middle. Padding pushes your key facts into that middle.
• It damages trust with people. Visitors who land on a stuffed page leave, and those who share it do not come back.

The alternative is one well-structured page per topic, with a clear answer, specific facts and sections for the real follow-up questions.

Which tactics are worth your time?

In rough order of return:

1. Fix access and accuracy: crawler access, content in HTML, consistent facts across your site and profiles.
2. Answer the questions buyers ask, directly and specifically, on the right pages.
3. Publish proof: case studies, pricing, security and methodology pages.
4. Earn independent mentions: press, expert contributions, reviews, directories.
5. Create original data that others want to cite.
6. Measure and iterate with a fixed set of buyer questions.

How to test which tactics actually move citation share is covered in what GEO tactics can improve AI citation share and how you test them.

How do you vet an agency's GEO tactics?

Ask for a written list of what they will do, and check it against the table above. Warning signs:

• Promises of guaranteed placement in ChatGPT or Google's AI answers.
• Networks of sites or "partner publications" they control.
• Review programmes that reward positive reviews.
• Hidden prompts, memory instructions or special content for AI crawlers.
• Sponsored placements they will not name or that carry no disclosure.
• Reluctance to explain methods because they are "proprietary".

A good agency will explain every tactic in plain terms and be comfortable with you publishing the list.

What might an illustrative case look like?

Consider an illustrative online lender whose previous agency had placed "best personal loans" articles on the sponsored sections of three news sites. For a few months, ChatGPT answers cited them. Then two of the host sites removed their sponsored sections after search penalties, and the citations vanished overnight. A journalist also wrote about the placements, and that article was cited in answers about the lender. The new team built a fair comparison page and a transparent rates and fees page, and published quarterly data on borrowing costs. Citations returned more slowly, but held.

What do people ask about GEO tactics?

Is gray-hat GEO illegal?

Usually not, but it can become so. Undisclosed paid endorsements and fake independent reviews breach consumer protection rules.

Will AI companies penalise black-hat tactics directly?

They defend against manipulation, and search indexes they rely on apply spam policies. The effect is often exclusion rather than a formal penalty.

Can competitors use black-hat tactics against us?

They can try, for example with fake negative content. Monitor mentions, report abuse to platforms and keep your own sources strong.

How do we know if a past agency used black-hat tactics?

Audit backlinks, sponsored placements, review sources and your site's code for hidden text or unusual crawler rules.