Personal Reputation Management: Remove, Correct & Suppress Results

The first practical decision is which surface contains the problem and which remedy can actually apply. Recruiters, investors, clients, landlords and reporters may check Google AI Overviews, ChatGPT and Perplexity alongside ordinary search results, so the audit must cover four working surfaces:

• Google search results: the page-one footprint anyone looking for you sees first, including images and news.
• Review platforms: Google Maps, Trustpilot, Yelp, Glassdoor, professional directories and marketplace profiles.
• AI answer engines: ChatGPT, Gemini, Perplexity, Claude, and Google's AI Overviews and AI Mode.
• Owned and earned media: profiles and content you control or can influence directly, plus coverage others publish about you.

Effective programs work all four surfaces together because the searcher's first touchpoint is often a third-party result rather than your own website.

Key Takeaway: Classify the surface and the problem before choosing a remedy. Skipping AI answers or review platforms leaves visible gaps untouched.

Step one is the audit. Search your name in a private browser window, then your name plus your city, employer, former employer, profession and words like "reviews," "lawsuit" or "complaint." Record every result on the first two pages, every image, and what each major AI assistant says when asked "Who is [your name]?" The free People Also Ask Explorer can help surface the name-plus-context questions people actually search. Classify each problem result as inaccurate, outdated, private, policy-violating, copyrighted, or accurate-but-unflattering; the classification decides which pathway below applies.

If you lead a company and your reputation is tied to it, the CEO/founder layer is covered in our Executive Reputation in the AI Answer Era guide. The rest of this guide is the individual playbook.

Negative Google results about your name are handled through one of five different outcomes: removal at the source, correction, deindexing (delisting from Google), a legal remedy, or suppression. They are not interchangeable, and picking the wrong one wastes months.

• Removal: the publisher or platform takes the content down. It disappears everywhere, not only from Google.
• Correction: the content stays up but is updated, amended or annotated so it is accurate.
• Deindexing / delisting: Google stops showing a URL for certain searches. The page still exists on the publisher's site and in other search engines.
• Legal remedy: a court order, or a legal request based on law such as defamation or privacy, that a publisher or Google reviews. This route typically requires qualified counsel.
• Suppression: nothing is removed. Accurate, relevant pages are built and strengthened so that, over time, they may appear above older results.

The table below maps each pathway to what qualifies, who makes the decision, and the official source. "Evidence class" marks how confident you can be: Documented (the platform publishes the rule), Observed (seen repeatedly in practice), Plausible (reasonable but not published) or Unproven.

How to use the table. Work top to bottom. Ask the publisher first, because a source-level fix is the only one that corrects the page everywhere. Use Google's personal-information tools for contact details and doxxing. Reserve legal pathways for genuine legal problems, with counsel. Use suppression only for accurate, lawful content that no pathway covers, and accept that a newsworthy, truthful article about your public conduct may remain visible whatever you do.

Key Takeaway: Removal, correction, deindexing, legal remedy and suppression are five different outcomes decided by different people. Match each problem result to the pathway it actually qualifies for, start with the publisher, and treat suppression as the fallback for accurate content, not a way to hide it.

Sometimes, when the review violates the platform's content policy, contains illegal content, or is subject to a valid legal order. "I don't like this review" is not a removal pathway. "This review is fake, off-topic, harassing or publishes my private information" can be.

This applies to reviews about you personally (as a doctor, lawyer, contractor, agent, landlord or consultant) as well as to reviews of a business you run. Google's Maps prohibited and restricted content policy lists categories that can be removed on policy grounds:

• Fake engagement: content not based on a real experience, including reviews from people who never dealt with you, and reviews that were paid for or incentivized.
• Off-topic content: political or social commentary, or personal rants unrelated to the experience.
• Harassment, hate speech and obscenity, including personal attacks and threats.
• Conflict of interest: for example, current or former employees reviewing an employer, or competitors reviewing each other.
• Personal information: reviews that publish phone numbers, addresses or other private details.

Reviews that contain false statements of fact (not opinion) may also be addressed through legal processes such as a court order. Whether a statement is defamatory is a legal question; speak with qualified counsel before alleging it.

What is not removable: an honest one-star review from a real customer or client describing a real experience. Even if it feels unfair, a genuine opinion is generally protected, and platform policy backs the reviewer. Anyone who promises to "remove all your bad reviews" is either overselling or about to use tactics that break platform rules.

How the documented process works:

1. Map to policy: match each flagged review to the specific policy category it violates, with screenshots and a short written rationale.
2. Report: use the platform's report tool. For a Google Business Profile, Google documents a report-and-review process in its review removal help page, including checking the status of a report.
3. Appeal once where available: if Google declines, the same help page documents a one-time appeal. Other platforms have their own processes.
4. Legal coordination: when a review may be unlawful, counsel can advise on a takedown request or a court order. Platforms review legal requests case-by-case.

Ethical guardrails. The FTC's rule on consumer reviews and testimonials, 16 CFR Part 465, prohibits fake reviews, buying positive or negative reviews, undisclosed insider reviews, and misrepresenting that a review site is independent, and it restricts using unfounded legal threats or intimidation to suppress reviews. In practice: no fake reviews, no undisclosed or incentivized reviews, no deceptive suppression, no impersonation or manipulation. These tactics can create regulatory and platform risk that is worse than the original review.

Key Takeaway: Negative reviews come down when they violate policy, contain illegal content, or are subject to a valid legal order. Honest negative opinions from real customers do not. The documented route is report, then a one-time appeal where available, then counsel for genuine legal issues. Never use fake or incentivized reviews.

Deindexing is available only when a result fits one of Google's documented categories (see the outcome definitions in the section above). This section covers who qualifies and what to file.

For a personal reputation problem, deindexing matters because it does not depend on persuading the publisher. It only applies when the content fits one of Google's documented categories, and Google makes the decision.

There are five primary deindexing pathways, each with its own standard:

1. Right to be forgotten: applicable EU and UK contexts. Under GDPR Article 17 and the parallel UK GDPR provision, people can ask search engines to delist results containing personal data that is inaccurate, inadequate, irrelevant or no longer relevant. Google reviews each request individually and balances privacy against the public interest in the information, including whether you have a public role.

2. Copyright: your own work. If a page reproduces your copyrighted photo, writing or video without authorization, Google's copyright removal form is the documented route. It only applies to real infringement; a critical article that does not copy your work is not a copyright problem.

3. Outdated Content tool. Google's outdated content removal process updates or drops a result when the underlying page has already been changed or removed at the source. It is the right route when a publisher has deleted or corrected a post but the old version still shows in Google.

4. Legal requests and court orders. Google reviews legal removal requests and court orders case-by-case through its legal help process. A court finding of defamation can support a request, but Google still assesses it; submission is not automatic removal. This pathway involves litigation or legal advice, so it should be coordinated with qualified counsel.

5. Personal information removal, including doxxing. Google's personal information removal policy covers contact details, government ID numbers, financial information, non-consensual explicit imagery and certain other categories. The Results about you tool lets you request removal of results showing your contact information and set alerts for new ones, where the feature is available.

The newsworthiness caveat. Google's policies state it may keep results that are in the public interest: for example, information about a person's professional conduct, public role or matters of public record. A truthful news article about you is usually not eligible for deindexing.

What deindexing does not do: it does not remove the content from the publisher's site, it does not remove it from other search engines (each has its own process), and it does not guarantee any change inside AI answers. See the AI section below.

This guide is not legal advice. Disputes involving defamation, privacy, harassment or data-protection rights may require a qualified lawyer in the relevant jurisdiction.

Key Takeaway: Deindexing pulls a URL out of Google's results without the publisher's cooperation, but only through documented pathways: right to be forgotten (applicable EU/UK contexts), copyright, Outdated Content, legal requests, and personal-information removal including Results about you. Google decides each request, and newsworthy information may stay visible.

Suppression is building and strengthening accurate, relevant pages about you so that, over time, they may appear above an older negative result. It is the fallback when a result is accurate, lawful and not eligible for any removal or deindexing pathway.

Plan on a minimum six-month program. Meaningful movement commonly takes 6–12 months or longer, depending on the strength of the negative URL, how common your name is, how competitive your name searches are, and the credible assets you start with. A thin blog post on a rarely searched name tends to move sooner than a major news article on a common name. Some results do not move at all. No one can guarantee a result will leave page one.

Anyone promising first-page suppression in 30 days is either inheriting a near-finished program or about to use spam techniques that tend to collapse and can damage your legitimate profiles.

Why it can work (Observed/Plausible, not guaranteed). Accurate, relevant, well-linked pages that clearly describe you can sometimes outrank older results over time, particularly when they come from credible sites and match what searchers are looking for. The underlying search mechanics are covered in our SEO and digital authority pillar guide.

As an example, not a requirement, a suppression program might build a portfolio of 12–25 accurate assets:

• Owned web properties: a personal website, a portfolio or speaker page, and a complete, accurate LinkedIn profile.
• Credible third-party profiles: professional association directories, licensing-board listings, alumni pages, conference-speaker profiles, and verified social profiles.
• Earned media coverage: bylines, podcast appearances, interviews and features in publications relevant to your field. This is where editorial PR becomes a reputation tool.
• Wikipedia and Wikidata: only when independent notability standards are genuinely met. Do not create or edit your own article in breach of Wikipedia's conflict-of-interest rules.
• Structured data on owned pages: Person schema that describes who you are and links to your verified profiles, so pages about you are clearly associated with each other.

Owned-profile and public-record repair. Before creating anything new, correct what already exists: outdated job titles, old bios on former employers' sites, mismatched photos, and directory entries with wrong details. Rebuilding an accurate public record draws on the same entity clarity and public evidence that Smart Money Media's Engineering Reference Authority methodology uses.

Why suppression is paired with AI correction. Moving a URL down Google's results does not stop ChatGPT or Perplexity from citing it. AI tools retrieve from the open web on their own terms and can surface a lower-ranking page in an answer. That's why suppression is paired with AI reputation correction, covered next.

Key Takeaway: Suppression is a minimum six-month program of building accurate, credible pages, with meaningful movement commonly taking 6–12 months or longer and no guaranteed outcome. It stacks owned, earned and profile assets against your name searches. It does not stop AI tools from citing the negative URL. That needs a parallel correction track.

AI reputation correction is the work of fixing inaccurate or misleading claims about you inside AI answer tools. ChatGPT, Gemini, Perplexity, Claude, and Google's AI Overviews and AI Mode. There is no single index to delist from, and no provider guarantees that a correction request will change future answers.

The four ways AI answers go wrong:

1. Hallucinated claims: the AI states something false with no grounding in any cited source. There is no upstream URL to fix, so platform reporting matters most here.
2. Cited-but-misframed claims: the AI cites a real source but presents it out of context, for example an old news story that no longer reflects current reality.
3. Outdated information: the answer reflects older material and misses developments such as a new role, a resolved dispute or a published correction. Training cutoffs vary by model and version.
4. Identity conflation: the AI blends you with someone who shares your name, attributing their history to you.

Documented correction channels (by provider):

• OpenAI (ChatGPT): Documented. OpenAI's privacy policy says that if output contains factually inaccurate information about you, you can request correction or removal through its privacy portal (privacy.openai.com) or by emailing dsar@openai.com, and that it will consider requests based on applicable law and technical capability. That is not a guarantee of correction.
• Anthropic (Claude): Documented. Anthropic's privacy policy describes privacy rights requests, including correction and deletion, subject to applicable law.
• Google (Gemini, AI Overviews): Documented feedback tools, outcome not guaranteed. Gemini and AI Overviews include in-product feedback, and the Gemini Apps Privacy Hub explains Google's privacy controls. Fixing the underlying Search results through the pathways above is often the more direct lever for AI Overviews (Plausible).
• Perplexity: Partly documented. Perplexity's privacy policy provides a privacy contact and correction requests for personal data; a dedicated process for correcting answers about a named person is Unproven; verify its current help pages before relying on one.

Correct the upstream sources. Many answers are assembled from pages retrieved at query time. Correcting or removing inaccurate source pages, and publishing clear, accurate information on credible pages about you, may improve browsing- and retrieval-based answers first (Observed/Plausible). Answers that rely on a model's training data may lag until later versions. For how different AI systems access and retrieve web content, see our LLM SEO guide.

What ChatGPT Memory does not do. ChatGPT Memory and Custom Instructions only affect your own conversations. Telling ChatGPT "that's wrong about me" in your account does not rewrite what other people see when they ask about you.

Measure it. Record answers from each major AI tool to a fixed set of prompts ("Who is [name]?", "Is [name] trustworthy?", "[name] [employer]") before you start, then re-run the same prompts monthly and log the date and engine. Without before-and-after records, AI reputation work cannot be evaluated.

Key Takeaway: AI answers go wrong in four ways: hallucination, misframing, outdated information and identity conflation. Correction runs through documented provider privacy channels, fixing upstream sources, and publishing accurate information, measured against a fixed monthly prompt set. No provider guarantees a change.

A reputation crisis is any sudden negative event (a viral review, a false blog post, a damaging news story, a coordinated attack) that is actively spreading and threatens material personal or professional harm.

Decisions made in the first 72 hours can strongly influence whether the event fades or becomes a lasting search result.

The protocol below is sequenced by hour because in a real crisis, the order of operations matters as much as the operations themselves.

Hour 0-6: Stabilize and assess.

• Capture a full record: screenshots, URLs, archive snapshots, share counts, and the likely original source.
• Identify the originating source and how it is spreading (who shared, when, to which audiences).
• Make an initial assessment of whether the content appears factually false (a possible legal question for counsel), policy-violating (a platform pathway may apply), or accurate-but-damaging (correction, context and suppression only).
• Agree on a single decision-maker for any statements.
• Pause scheduled posts, announcements or outreach that could collide with the event.

Hour 6-24: Containment.

• If the content violates platform policy, file the appropriate report promptly.
• If you believe it is defamatory or violates your privacy, speak with qualified counsel before sending any legal threat or publicly alleging defamation or a privacy violation. Unfounded threats can backfire and draw more attention.
• For coordinated attacks, document the originating accounts; many "viral" events trace back to a handful of sources that can be reported individually.
• Prepare a short, factual holding statement if the event is visible to clients, employers or press.

Hour 24-72: Context and accurate information.

• Publish a factual, calm response or correction on your own credible profiles where appropriate.
• If there is a fair story to tell, a balanced, on-record article from a credible reporter can add context, and may over time appear alongside or above the original. There is no guarantee on search displacement.
• Start monitoring AI answers for citations of the originating source.
• Brief the people who matter most (employer, clients, board or family) before they find it themselves.

What not to do in a crisis. Do not respond publicly to anonymous attackers (it can amplify the source). Do not threaten litigation publicly without legal advice. Do not buy reviews, fake engagement or use services that violate platform terms.

Key Takeaway: The first 72 hours follow a sequenced protocol: stabilize, contain, add context. The most damaging mistakes tend to happen in the first 24 hours: amplifying anonymous attackers, threatening legal action without counsel, or buying fake reviews to dilute the negative ones.

If you are an executive or founder, the same playbook applies, but four asymmetries change the priorities. First, the audience differs: investors, board members, recruits and partners. Second, name searches cluster around context ("[name] [company]," "[name] previous role"), so the negative result that matters is often not on your name alone. Third, AI tools tend to describe people in longer narrative paragraphs than products, which creates more places for outdated claims to land. Fourth, personal-data pathways may be available to an individual that are not available to a company, so audit them separately.

A Knowledge Panel built on accurate, verifiable sources can help anchor your name search for people who meet notability thresholds (Plausible/Observed, not guaranteed. Google decides whether a panel appears and what it shows).

For the full CEO and founder treatment (authority building and how executives are described in AI answers) read our Executive Reputation in the AI Answer Era guide.

Key Takeaway: Executives face four asymmetries: audience, context-based name searches, longer AI narratives and separate personal-data pathways. Use this playbook for the fundamentals and the executive guide for depth.

Ethical personal reputation management corrects the record; it does not hide the truth. These boundaries apply to every step in this guide:

• Remove only when eligible: platform policy, Google's personal-information or legal policies, the right to be forgotten where it applies, genuine copyright infringement, or a valid court order.
• Correct inaccuracies at the source first: ask the publisher or platform, then use the search and AI provider channels.
• Suppression means building an accurate, credible public record, never hiding wrongdoing from people with a legitimate interest in it. Newsworthy, truthful information may stay visible.
• No fake or incentivized reviews, no review gating, and no paying for removal of honest reviews (Google Maps policy; FTC 16 CFR Part 465).
• No impersonation, sockpuppets, private blog networks (PBNs) or fake legal notices.
• Legal counsel for defamation, privacy and data-protection disputes. This guide is not legal advice.
• No vendor can guarantee removal, deindexing, rankings or AI-answer changes. Platforms and search engines make those decisions.

Key Takeaway: Remove only what is eligible, correct at the source first, build an accurate record rather than hiding the truth, never fake reviews or identities, and bring in qualified counsel for legal disputes.

Measure reputation work with dated snapshots, not impressions. Track the same items on the same schedule (monthly is typical), so changes can be compared over a minimum six-month horizon:

• Name-search snapshots: screenshots of page one and two for your name and each name-plus-context search, dated, from a private browser window.
• Results about you alerts: where available, new results showing your contact details.
• A fixed AI prompt set: the same prompts, recorded by engine and date.
• Review sentiment: rating, volume and the themes in recent reviews.
• Profile and entity accuracy: whether your bios, titles, photos and directory entries match across sites.
• Visibility of credible sources: how many owned or credible third-party pages appear for your name.

To see how AI tools currently describe you and which sources they draw on, run the free AI Visibility Audit and repeat it on the same schedule.

Keep the limits in view. Visibility changes are not the same as reputation outcomes. A negative result moving down, or more mentions and citations of accurate pages, does not by itself prove that people trust you more or that opportunities or revenue will follow. Judge progress over months, not days.

Key Takeaway: Track dated name-search snapshots, alerts, a fixed AI prompt set, review sentiment, profile accuracy and credible-source visibility. Visibility is a signal, not proof of trust.

The reputation management industry includes vendors who oversell removal capabilities or use tactics that break platform rules. Vet carefully, whoever you choose.

Red flags that should end the conversation:

• "We can remove any negative review." Honest opinions from real customers are not removable through legitimate pathways.
• "We guarantee first-page suppression in 30 days." Suppression is a minimum six-month program with meaningful movement commonly taking 6–12 months or longer. Fast promises usually mean spam techniques.
• "We have a relationship with Google to remove results." Google does not give removal authority to vendors. Requests go through documented public channels anyone can use.
• "We'll bury your results with hundreds of microsites." Private blog networks and link farms can get legitimate profiles penalized.
• No measurement. If there are no dated before-and-after records for your name searches and AI prompts, the work cannot be evaluated.

Green flags:

• Honest scoping that separates removable, correctable and suppressible results.
• A named pathway and policy or legal basis for each flagged URL or review.
• Suppression built on real, accurate profiles and earned coverage.
• An AI correction track using documented provider channels and a fixed prompt set.
• Transparent reporting on every request filed, granted or denied.

The question that filters most. Ask: "What specifically can you not do, and what should I expect to remain visible after six months?" Honest answers signal a disciplined operator.

Key Takeaway: Vet on what a vendor refuses to do, what it cannot do, and how it measures progress, not on the size of its promises.

For the broader trust foundation, see our brand credibility pillar, and for how AI search quotes sources, our AEO and GEO guide.

To see what an honest scope looks like for your situation, take the free 5-question reputation risk assessment, start with an AI Visibility Audit, or review our reputation management service.