Discussion starter: an AI search question, posted with an answer from Niraj Raut to open the thread. If you have dealt with this on a site, reply with what you saw, especially where it differs.
- AI Search
Where do you draw the line between legitimate AI visibility work and manipulation, such as paying for placement in third-party best-of lists?
Disclosure: Niraj Raut, who posted this answer, runs the SEO consultancy linked at the end of it.
1 reply
-
adminAdmin
I draw the line at disclosure and truthfulness. Work that helps accurate information about you reach the sources assistants use is legitimate: earning reviews and coverage, briefing reviewers, publishing clear product facts, correcting errors, running a disclosed affiliate programme. Work that manufactures the appearance of independent endorsement is manipulation: paying for a top spot in a “best of” list that presents itself as editorial, planting unlabelled posts in communities, running review sites you secretly control, or hiding instructions for language models in your pages. The first kind survives platform changes and scrutiny. The second kind is exposed to spam policies, consumer protection law and reputational damage, and it tends to work only until someone notices.
What the platforms and regulators say
Google’s May 2026 guide to optimising for its generative AI features says that seeking inauthentic mentions is not helpful, and its spam policies already cover the main mechanisms: link schemes that include paid links without qualification, and site reputation abuse, where third-party content is placed on a host site mainly to exploit its ranking signals. A paid “best of” list published on a high-authority domain with no real editorial involvement can fall under the second policy, and outside the EEA that can still mean a manual action on the host section.
Consumer law is the other constraint. Advertising regulators in the UK, US, Australia and the EU generally require paid placements to be identifiable as advertising, and the US Federal Trade Commission’s 2024 rule on consumer reviews and testimonials targets, among other things, company-controlled review sites presented as independent. If an invoice determines a ranking, a reader who cannot see that is being misled, whatever the SEO outcome.
A working spectrum
Activity Where I put it Why Giving reviewers access, data and briefings, with no control over the verdict Legitimate Standard PR; the reviewer’s judgement stays independent Correcting factual errors in third-party lists and knowledge sources Legitimate Improves accuracy for everyone Publishing your own comparison pages that say they are yours and treat competitors fairly Legitimate Readers know whose view it is Affiliate programmes with clear disclosure Legitimate, with care Disclosure lets readers weigh the incentive; problems arise when commission quietly sets the order Paying for inclusion in a list that is clearly labelled as sponsored Grey Disclosed, but many readers skim past labels, and the list’s value as a recommendation drops Staff posting in communities without saying who they work for Manipulation Deceives other users and breaks most community rules Paying for a top position in a list presented as independent Manipulation Misleads readers, risks spam action on the host, and may breach advertising rules Running “independent” review sites you control Manipulation A clear consumer protection risk as well as a trust risk Hidden text or instructions aimed at language models Manipulation Cloaking in intent; AI providers treat prompt injection as a security problem Three questions I use for grey cases
- Would the reader change their view if they saw the commercial arrangement? If yes, and they cannot see it, it is manipulation.
- Does the tactic add accurate information, or only an endorsement signal? Supplying specs, pricing and test units adds information. Paying for a rank adds only a signal.
- Would it still work if every platform could see exactly what you did? Tactics that depend on not being noticed are not durable, because platforms, journalists and competitors are actively looking.
Why the manipulative end is fragile for AI visibility specifically
Paid list placements work today because some assistants retrieve pages that rank well and summarise them, and a list on a strong domain ranks well. That chain has several weak links. Search engines can reclassify or demote the host section. AI providers can learn to discount self-serving or sponsored lists as source types, in the same way that search engines learned to discount paid links. And journalists and community moderators increasingly write about brands that game AI answers, which turns a visibility tactic into a reputation story. None of that has to happen quickly for the investment to be a poor one; it only has to happen before the placement has paid back.
Hypothetical example:
A software company pays several publishers for top placement in “best tools” lists without disclosure. For a few months assistants often name the company for category prompts. Then a core update and a site reputation review demote two of the host sections, and a community thread compiles the paid lists with screenshots. The assistants stop citing the lists, and the thread starts appearing for the brand’s name. The company would have been better served by the harder route: getting a small number of independent reviewers to test the product.
When competitors are already doing it
The hardest version of this question is not “should we?” but “they are, and it is working, so why shouldn’t we?” A few things help in that conversation.
- Document it. Keep dated screenshots of the paid placements, the assistant answers that cite them and any missing disclosures. If the tactic later becomes a spam or regulatory issue, you will want that record, and it is useful for a spam report now.
- Report where there is a channel for it. Google accepts spam reports, and advertising regulators take complaints about undisclosed paid endorsements. Reports rarely produce fast results, but they are not wasted.
- Compete on the sources the paid lists cannot buy. Independent reviewers, analyst coverage, real customer reviews and community reputation are harder to fake and are exactly the sources that assistants and search engines are likely to lean on more if they start discounting sponsored lists.
- Put a number on the downside. For leadership, frame the risk as a cost: a manual action on a partner site, a news story about fake endorsements, or a regulator’s letter. The expected cost of those outcomes often outweighs a few months of extra mentions.
It is frustrating to lose ground to a tactic you will not use. It is more frustrating to explain to a board why the brand is in a story about paid AI rankings.
What legitimate AI visibility work looks like in practice
- Make product facts, pricing structure, integrations and limitations easy to find on your own site, in text.
- Identify the independent reviewers, analysts and communities that assistants cite for your category, and give them real access.
- Correct errors in third-party profiles, knowledge sources and comparison sites.
- Earn reviews from real customers on the platforms your buyers use, without incentives that break those platforms’ rules.
- If you use affiliates or sponsorships, require clear disclosure in your contracts, and do not buy rank position.
- Track whether assistants describe you accurately, not just whether they mention you.
My line, stated simply: influence what is true and how easily it can be found, and never pay to change what an apparently independent source says about you without the reader knowing. That line also happens to be where the durable results are.
Need help with this on your own site?
Niraj Raut works with businesses in Nepal, Australia, the UK, Europe and the US on technical SEO, ecommerce SEO, local SEO and AI search.
Add a reply Cancel reply
Replies are open to approved members. Log in or apply to join.