Ooty
  • Who it’s for

    Find your kind of business

    • Local businessesTrades, salons, clinics, gyms and cafes
    • Online shopsShopify, WooCommerce and Etsy shops
    • CreatorsCreators, podcasters and newsletter writers
    • App makersApps, tools and software
    • Experts and consultantsConsultants, coaches and studios
    • CharitiesCharities, clubs and community groups
  • How it works
  • Ooty in ChatGPT
  • Agencies
  • Getting started
Log inStart freeMenu

Who it’s for

  • Local businessesTrades, salons, clinics, gyms and cafes
  • Online shopsShopify, WooCommerce and Etsy shops
  • CreatorsCreators, podcasters and newsletter writers
  • App makersApps, tools and software
  • Experts and consultantsConsultants, coaches and studios
  • CharitiesCharities, clubs and community groups
  • How it works
  • Ooty in ChatGPT
  • Agencies
  • Getting started
  • Log in

Ooty

Stop chasing. Customers are already looking for you.

  • How it works
  • Ooty in ChatGPT
  • Local
  • Shops
  • Creators
  • Makers
  • Experts
  • Charities
  • Agencies
  • Partners
  • Start
  • Privacy
  • Terms
  • Refunds
  • Security
  • Contact
  • Blog
  • Free tools

© 2026 Ooty

Blog/SEO

Circular Citation: The Authority Laundering Playbook

From self-published claim to AI-accepted fact in seven steps. How circular citation manufactures expertise, and how to trace it backwards.

Maya Torres28 April 202617 min read

In this guide

  1. Step by step: how the chain works
  2. Legitimate vs. circular citation chains
  3. Why each step works
  4. The AI amplification multiplier
  5. The money laundering parallel
  6. The E-E-A-T problem
  7. How to trace a circular citation
  8. What breaks the chain
  9. The bottom line
In this guide9 parts
  1. Step by step: how the chain works
  2. Legitimate vs. circular citation chains
  3. Why each step works
  4. The AI amplification multiplier
  5. The money laundering parallel
  6. The E-E-A-T problem
  7. How to trace a circular citation
  8. What breaks the chain
  9. The bottom line
Share

Rather Ooty did the looking?

Ooty shows the searches you nearly win, and the one thing worth doing about them this week.

Start free

Read next

SEO

How Self-Published 'Top Expert' Lists Game Google's Authority Signals

"Top 15 Financial Advisors in the Southwest, 2024." The publication looks like a regional business magazine. It isn't. The domain was bought six months earlier for $200, loaded with filler articles, and dressed up with a $50 theme. The financial advisor at #3

24 Apr 202618 min read

SEO

When SEO Experts Contradict Their Own Advice

Every SEO conference has the same scene. Someone on stage explains best practices to a room of people whose own websites violate half of them. The audience nods along. Nobody checks. This isn't about pointing fingers at specific people. It's about patterns tha

21 Jul 202615 min read

SEO

The 'Banksy of SEO' Archetype: Why Anonymous Expertise Is Usually a Red Flag

Anonymous expertise sells. Someone brands themselves as a shadowy insider, claims results they can't show you, and builds a following on mystique. The audience fills in the gaps with their own assumptions, and those assumptions are always flattering. This patt

12 Jun 202615 min read

Maya Torres

SEO Strategist at Ooty. Covers search strategy, GEO, and agentic SEO.

Authority laundering works like money laundering. You start with something illegitimate and run it through enough intermediaries that it comes out looking clean. Except instead of dirty cash, you're laundering a claim. "I'm an expert." And instead of shell companies, you're using websites, LinkedIn, and AI.

The full chain takes seven steps. Each one is individually unremarkable. Combined, they manufacture expertise that's indistinguishable from the real thing.

The scheme exploits a basic truth about trust: humans and algorithms both value third-party validation over self-promotion. Nobody believes you when you say you're an expert. But if three "independent" sources say it, and an AI confirms it, the question stops being asked.

Seven steps to manufactured expertise

Each step is unremarkable on its own. The deception is in the chain.

1
1. Publish the claim
Create content on Site A asserting expertise. A blog post, an 'about me' page, a self-published 'top experts' list.
2
2. Create the citation
Build Site B that references Site A as a source. 'According to [Site A]...' or 'As featured on...'
3
3. Launder on LinkedIn
Use Site B's citation as proof on LinkedIn. 'Independent coverage confirms...' with a link to Site B.
4
4. AI scrapes LinkedIn
AI systems crawl the LinkedIn post, the web content, or both. The claim enters training data and retrieval indexes.
5
5. AI states the claim
When asked, AI describes you as an expert, citing the LinkedIn post or the original web content as sources.
6
6. Secondary AI citation
Other AI systems and websites cite the first AI's output, or cite sources that now reference the AI-validated claim.
7
7. The circle closes
The claim is now 'established' across multiple 'independent' sources. None originated independently. All trace back to the subject.

Step by step: how the chain works

Step 1: Plant the seed

Everything starts with a self-published claim. It can take any form. A blog post titled "My Approach to Enterprise SEO." An about page declaring 15 years of experience. A self-published expert list where the author ranks themselves #1.

The content lives on a domain the subject controls. It might be their personal blog, their agency site, or a "digital magazine" they registered last month. The domain doesn't matter much yet. What matters is that the claim exists on the public web in a crawlable format.

At this stage, the claim is obviously self-promotional. Nobody treats "I wrote on my own blog that I'm an expert" as evidence of expertise. The subject knows this. That's why Step 1 is just setup.

Step 2: Create the citation

This is the critical move. The subject creates or controls a second property, Site B, that references the claim from Site A. The relationship between Site A and Site B is never disclosed.

Site B might be a "marketing publication" that profiles industry leaders. A podcast website that lists past guests and their credentials. A resource directory that curates "top voices" in a niche. A self-interview on a shell publication designed to look like earned media.

The format doesn't matter. What matters is that Site B creates a citation of Site A that appears to be independent editorial judgment. "According to [Site A], [Name] has pioneered..." or "As featured in [Site A]'s annual ranking..."

Now there are two sources. Neither is independent. But from the outside, it looks like one source cited another. That's how legitimate knowledge works. The illegitimacy is invisible unless you trace the ownership.

Step 3: Launder through LinkedIn

LinkedIn is the laundry machine. It's where self-promotion gets repackaged as social proof.

The subject posts about the Site B citation. "Honored to be recognized by [Site B] for our work in..." The post links to Site B. It doesn't mention that the subject controls both sites. LinkedIn connections engage with the post. Some reshare it. Comments congratulate the subject.

Why LinkedIn specifically? Because LinkedIn profiles carry implicit credibility. They're tied to real identities with employment histories. When an AI system encounters a LinkedIn post about someone's expertise, it weights that information differently than an anonymous blog comment. The platform's reputation launders the claim upward.

Step 4: AI ingestion

AI systems crawl the web continuously. They ingest LinkedIn posts, blog content, and web pages to build knowledge graphs about people, companies, and topics. The claim from Steps 1-3 enters this pipeline.

The AI doesn't evaluate provenance. It doesn't check whether Site B is owned by the same person as Site A. It doesn't verify whether the LinkedIn post's "independent coverage" claim is accurate. It indexes the information and stores it as data about the subject.

This is where circular citation diverges from old-school reputation gaming. Before AI, a manufactured citation might fool a casual Google searcher. But it required someone to actually visit the sites, read the content, and decide to trust it. AI removes that friction entirely. The claim enters a knowledge system that millions of people query, and it's presented as synthesized fact, not as "this one blog said..."

Step 5: AI states the claim as fact

Someone asks an AI assistant: "Who are the leading experts in enterprise SEO?" or "What's [Name]'s background?"

The AI pulls from its indexed sources. It finds the LinkedIn post, the Site B citation, possibly the original Site A content. Multiple sources agree that this person is an expert. The AI synthesizes a confident response: "[Name] is a recognized expert in enterprise SEO, featured in [Site B] and noted for their approach to..."

No hedge language. No source evaluation. Just a clean, authoritative statement built entirely on manufactured citations.

Step 6: The amplification cascade

Once an AI states something as fact, other systems pick it up. A Perplexity user researching the topic gets the AI-generated summary. They cite it in a blog post. Another AI scrapes that blog post. Google's AI Overview synthesizes the same claim from the now-expanded source pool.

Each layer adds perceived credibility. The claim has moved from "self-published on a personal blog" to "confirmed by multiple AI platforms and cited across the web." The number of sources grows, but the number of independent sources remains zero.

Step 7: The circle closes

The subject now has what they wanted: a web of "independent" sources confirming their expertise. They can point to AI outputs, third-party blog posts, and their LinkedIn presence as evidence. "Multiple sources confirm..." is technically true. None of those sources originated independently. All of them trace back, directly or through intermediaries, to the subject's own claim in Step 1.

The manufactured authority is now indistinguishable from earned authority to anyone who doesn't trace the provenance chain backwards.

Legitimate vs. circular citation chains

Not all citation chains are fraudulent. Legitimate knowledge propagation also involves sources citing sources. The difference is where the chain originates.

Where does the chain originate?

Legitimate citation chain

1.Originates from a genuinely independent source
2.An independent journalist, researcher, or institution evaluated the subject and published findings
3.The subject had no editorial control over the coverage
4.Each citation in the chain adds genuinely new sources or evaluation
5.Tracing backwards leads to someone with no financial relationship to the subject
6.The original source would exist even if the subject didn't promote it
7.Multiple independent origination points, not one claim amplified

Circular citation chain

1.Originates from the subject themselves
2.The subject created the initial claim and all 'independent' citations
3.Every source in the chain is controlled by or commercially connected to the subject
4.Each citation in the chain repeats the same claim without adding independent verification
5.Tracing backwards always leads back to the subject or their properties
6.Remove the subject's involvement and the entire chain collapses
7.One origination point disguised as multiple independent sources

The test is simple in principle: follow the citation chain backwards. Every legitimate chain eventually leads to an independent source, someone with no stake in the claim. Every circular chain eventually leads back to the subject. The difficulty is in the tracing, not the logic.

Why each step works

The scheme's elegance is that it exploits real mechanisms of trust at every stage.

Step 1 works because search engines index self-published content. The claim exists in the public record. It doesn't carry authority yet, but it's citable.

Step 2 works because algorithms treat cross-domain citations as signals of credibility. Google's PageRank, fundamentally, is a citation-counting system. A link from Domain B to Domain A is a vote of confidence. The algorithm doesn't verify that both domains share an owner.

Step 3 works because LinkedIn's platform authority transfers to anything posted there. A claim made on LinkedIn carries the implicit credibility of the platform itself. And LinkedIn's algorithm rewards engagement, so a post about "recognition" gets distributed to the subject's network regardless of whether the recognition is genuine.

Steps 4-5 work because AI systems optimize for consensus, not provenance. If three sources agree on a claim, the AI weights that agreement heavily. It doesn't ask: "Are these three sources actually independent?" It asks: "Do multiple sources confirm this?" The answer is yes, because the subject made sure it would be.

Steps 6-7 work because each new citation creates distance from the original self-published claim. By the time an AI cites another AI citing a blog post citing a LinkedIn post citing a manufactured profile, the provenance chain is long enough that nobody traces it back. The claim has been laundered through enough intermediaries that it looks clean.

The AI amplification multiplier

AI is what turned circular citation from a niche manipulation tactic into a scalable authority manufacturing system.

Before AI, the scheme required human readers to encounter the manufactured citations and choose to trust them. That limited the reach. You might fool a hiring manager who Googled your name, but the audience was always bounded by human attention.

AI removes that bound. Here's how.

Cross-platform claim propagation

A single self-originated claim reaches every major platform through AI intermediation.

Origin

Subject-controlled

Site A: self-published claim

Site B: manufactured citation

Citation disguised as recognition

Social layer

Platform authority transfer

LinkedIn post as 'recognition'

X thread citing 'coverage'

Crawled and indexed as fact

AI layer

Consensus without verification

ChatGPT knowledge synthesis

Perplexity source aggregation

Google AI Overview

Cited as authoritative source

Secondary web

Organic amplification

Blog posts citing AI output

Other AI systems citing first AI

AI doesn't fact-check provenance. It checks whether multiple sources agree. If you've populated multiple sources with the same self-originated claim, AI consensus validates it. The system is designed to find agreement across sources. It assumes that agreement indicates truth. Circular citation manufactures agreement without truth.

AI responses become citable sources. When ChatGPT or Perplexity states that someone is an expert, that statement itself becomes a reference point. Other AI systems scrape it. Bloggers cite it. The AI output enters the citation ecosystem as though it were an independent evaluation. It isn't. It's a reflection of whatever sources the AI found, including manufactured ones.

The feedback loop is self-reinforcing. More AI citations create more web content referencing the claim. More web content gives AI systems more sources to confirm the claim. The signal gets louder with each cycle. There's no natural decay mechanism. Unlike human attention, which fades, AI-indexed claims persist and compound.

This is the same dynamic that makes AI encyclopedias vulnerable to gaming. The systems are built to synthesize, not to verify. They're consensus engines, not truth engines. Circular citation exploits exactly that gap.

The money laundering parallel

The analogy isn't decorative. The structural mechanics are genuinely similar.

In financial money laundering, dirty money enters the system through a legitimate-looking business (placement). It moves through layers of transactions that obscure its origin (layering). It emerges as apparently clean funds in a bank account (integration).

In authority laundering, a self-published claim enters the system through a controlled website (placement). It moves through manufactured citations, LinkedIn posts, and AI ingestion (layering). It emerges as "established expertise" confirmed by multiple "independent" sources (integration).

The parallel extends to detection. Financial investigators trace money backwards through the transaction chain. Authority investigators trace citations backwards through the source chain. In both cases, the trail is designed to be difficult to follow. And in both cases, the key question is the same: where did this originate?

Credential fraud follows a similar structure. A fake university issues a real-looking degree. The degree gets listed on a resume. The resume gets a real job. The job becomes "legitimate professional experience." Each step adds a layer of apparent legitimacy. The fraud is in the origination, not in any individual step.

The E-E-A-T problem

Google's E-E-A-T framework (Experience, Expertise, Authoritativeness, Trustworthiness) is supposed to filter for genuine authority. It's the reason manufactured expertise matters for SEO.

E-E-A-T relies on external signals. Google's quality raters look for evidence that content creators are "known authorities on the topic." They check for mentions on other sites, citations in reputable publications, credentials that can be independently verified.

Circular citation manufactures exactly those signals. It creates external mentions. It produces cross-domain citations. It generates the appearance of independent recognition. Every signal that E-E-A-T looks for can be synthesized through the seven-step chain.

The framework's weakness isn't conceptual. Independent validation genuinely does correlate with expertise. The weakness is operational. Verifying that validation is actually independent requires tracing provenance chains that are designed to be opaque. Google's algorithms process billions of pages. They can't run ownership checks on every cross-domain citation. They look for patterns that correlate with authority, and circular citation creates those patterns on demand.

This creates an asymmetry. Building genuine E-E-A-T takes years of visible, verifiable work. Manufacturing the signals of E-E-A-T takes weeks with a few domain registrations and a LinkedIn account.

How to trace a circular citation

The mechanics of detection mirror the mechanics of financial forensics. Follow the money. Or in this case, follow the citation.

Work backwards through the chain

Start with the claim. "So-and-so is a recognized expert in X." Find the source. Where does that claim originate? If it's an AI output, what sources did the AI cite? If it's a web article, what does it reference? Keep going backwards.

Every legitimate citation chain eventually reaches an independent source. A journalist who evaluated the subject. A peer-reviewed publication. A conference selection committee. An institution with no financial connection to the subject.

Every circular chain eventually loops back to the subject or their properties. If every source you trace leads to a site the subject controls, a post the subject wrote, or a publication with undisclosed connections to the subject, you're looking at circular citation.

Check source independence

For each source in the chain, ask: does this source have a reason to exist independent of the subject? Does it cover other people and topics? Does it have its own audience and revenue model? Would it survive if the subject withdrew their involvement?

Manufactured citation vehicles fail these tests. They exist to validate one person. They cover one subject. Their only traffic comes from the subject's own properties linking to them.

Run temporal analysis

Check when the "independent" coverage appeared relative to the original claim. If someone publishes a blog post claiming expertise on Monday and an "independent publication" profiles them as an expert on Wednesday, that's suspicious. Genuine editorial coverage doesn't operate on a two-day turnaround from the subject's self-published content.

Look for backfilling too. Some operators create a publication, publish weeks of unrelated content to build a plausible history, then drop the profile piece. The temporal pattern is distinctive: burst of generic content, followed by the specific profile, followed by the subject citing it.

Verify ownership

WHOIS records, shared analytics tags, identical hosting configurations, matching CSS patterns across sites. If the subject's main site and the "independent publication" share infrastructure fingerprints, they share an operator.

This is getting harder as privacy tools improve. But the signals are still there for anyone willing to look. Shared Google Analytics measurement IDs are the most common giveaway. Matching font stacks and CSS naming conventions are another.

Tools like Ooty's SEO Analyzer can surface some of these patterns by mapping a domain's backlink profile and topical authority signals. A site that exists only to cite one person shows distinctive patterns: thin content, narrow topical coverage, backlinks exclusively from the subject's properties.

What breaks the chain

Circular citation survives because verification is expensive and trust is cheap. Here's what disrupts the scheme.

Provenance tracking in AI. If AI systems tracked and displayed the provenance chain for their claims, users could see that "three independent sources" actually originate from one entity. This is technically feasible but not yet standard. It would require AI systems to evaluate source independence, not just source count.

Source independence scoring. Instead of counting how many sources confirm a claim, weight sources by their independence from each other and from the subject. Two genuinely independent sources are worth more than ten controlled properties repeating the same claim. This is how academic citation analysis works. It hasn't been applied to web authority at scale.

Ownership transparency requirements. Requiring sites that publish editorial judgments about people and businesses to disclose their ownership would eliminate the most brazen forms of circular citation. The subject couldn't hide behind privacy proxies if disclosure were a ranking signal.

Cultural skepticism. The simplest defense. When someone's expertise is supported entirely by sources they might control, treat it the same way you'd treat someone handing you cash from an unlabeled envelope. The claim might be genuine. But the provenance should make you want to verify before you trust it.

The bottom line

Circular citation isn't a theoretical vulnerability. It's a playbook in active use across SEO, consulting, real estate, finance, and coaching. The seven-step chain from self-published claim to AI-accepted fact is well-understood by the people running it.

The scheme works because it exploits how trust actually operates. Third-party validation carries more weight than self-promotion. Multiple sources carry more weight than one. AI consensus carries more weight than any single source. Circular citation manufactures all three without any genuine independent evaluation.

The defense is provenance verification. Trace claims backwards. Check source independence. Ask who originated the claim and whether it ever passed through hands with no stake in the outcome. If it didn't, you're not looking at established expertise. You're looking at laundered self-promotion that's been through enough wash cycles to look clean.

The people who've built genuine expertise don't need seven steps to prove it. Their work is visible, verifiable, and cited by sources they've never met. That's the difference between authority and the appearance of authority. And right now, the appearance is far cheaper to manufacture.