1. Why the vocabulary is a mess

People use five labels as if they were synonyms. They are not.

Two academic distinctions matter.

Negative vs. dirty campaigning

Political communication research separates negative campaigning (critical, civil, issue- or record-based attack) from dirty campaigning (incivility, defamation, below-the-belt methods). Treating every attack as "just negative campaigning" collapses a line that courts and voters still draw.

White / gray / black communication

A useful ethics map, borrowed from propaganda studies and later applied to PR:

• White: source is identified; claims are intended to be checkable; persuasion is overt.
• Gray: source or motive is partly hidden; framing is selective; legality may hold while ethics erode.
• Black: source is concealed or falsely attributed; the aim is to deceive about identity, motive, or facts.

Legitimate public relations lives in white, sometimes brushes gray (ghostwritten op-eds, unnamed "people familiar with"), and is not black. Once the sponsor is hidden and the content is built to injure, you have left the profession.

Russian and Eastern European literature is the origin of the modern "black PR" label. Alena Ledeneva and later policy reports describe chernyi piar as a bundle of informal practices that blur legal and illegal means of ruining a rival: kompromat, paid-for coverage, identity confusion in elections, and reputation destruction as a political technology. The term then migrated into commercial usage worldwide.

A 2023 systematic look at "Black PR" and "Dark PR" in Google and Google Scholar found that the two labels are used almost interchangeably, with "dark PR" appearing later and more often attached to digital tactics. Both describe professionalized disinformation aimed at discrediting an adversary.

Working definition for this brief

A smear / dark-PR campaign is a planned, multi-channel effort to degrade a target's public standing, in which at least one of the following is true: the sponsor is concealed; material facts are fabricated or systematically distorted; amplification is inauthentic; or the operation is designed so the target cannot answer on equal terms.

Negative publicity is the visible outcome (coverage, search results, social narrative). It can be produced by a smear, by journalism, by a real scandal, or by all three at once.

That last sentence is why agencies get this wrong. They treat "we are in the press" as the campaign. The campaign is the system that put them there and keeps them there.

2. What is being attacked (the asset)

Reputation is not a feeling. Operationally it is a bundle of assets:

• Search identity — page-one results for the name or brand.
• AI identity — what large language models and answer engines say when asked "who is X / is X legitimate."
• Media identity — the stock story reporters start from.
• Peer identity — what employees, investors, counterparties, and other agencies repeat privately.
• Institutional identity — regulator, bank, insurer, and platform-trust posture.
• Emotional identity — the one-line moral verdict ("predator," "fraud," "hypocrite," "incompetent").

A mature campaign does not need to "win" all six. It needs to contaminate the two or three that decide the next deal, hire, election, or partnership. Empirical work on executive search results after negative press finds the first week does disproportionate damage: in one 100-crisis study, negative items reached page one within three to five days in 82% of cases, and a large share of prior positive URLs were displaced within a week. Time later recovers some ground, then flattens. The scar often remains.

That is why smear logic targets search and narrative lock-in, not just a single article.

3. A taxonomy of campaigns (by structure, not by recipe)

Do not classify campaigns by slogan. Classify them by sponsor visibility, truth relationship, channel mix, and objective.

3.1 By sponsor visibility

• Overt adversarial campaign. A named rival, union, activist coalition, or political opponent attacks on the record. Legal exposure is higher for the sponsor; so is accountability. This can still be a smear if the content is false.
• Proxy campaign. A law firm, NGO, "concerned citizens" vehicle, or friendly outlet carries the message. The economic beneficiary is one step removed.
• Covert campaign. The commissioning party is designed to stay invisible. This is the core of what trade writers call dark PR. Concealment itself is often the most legally and ethically significant fact.
• Opportunistic pile-on. No single sponsor. A real incident is hijacked by unrelated actors. Looks like a campaign; is often a cascade.

3.2 By relationship to truth

• True-and-weaponized. The facts are accurate; selection, timing, and framing are the weapon. Journalism and smear can look identical from one screenshot. Motive, omission, and coordination are what distinguish them.
• Half-true / context-stripped. A real document, quote, or photo is cut so the remaining shard implies a different crime.
• False. Fabricated events, forged documents, impersonation, synthetic media.
• Unfalsifiable. Innuendo, "people are saying," guilt by adjacency, ritual questions ("why won't they deny…").

True-and-weaponized material is often harder to kill than a lie. Courts and platforms give more protection to accurate statements. The defense is context, not deletion.

3.3 By objective

Researchers and practitioners see recurring end-states:

• Character assassination — make the person the story, so policy or product becomes undiscussable.
• Deal disruption — scare a buyer, investor, regulator, or employer during a window (IPO, merger, confirmation, fundraising).
• Market effect — short-and-distort and related information attacks on listed names (securities law territory).
• Cancellation / isolation — make association costly for third parties.
• Pre-emption — smear first so the target's future accusation looks like revenge.
• DARVO pattern — deny, attack the accuser, reverse victim and offender. Named in trauma research by Jennifer Freyd (1997); now widely observed in institutional and political fights.

3.4 By theatre

Campaigns occupy stacked theatres. A single operation may use several. Naming the theatres is enough for education; specifying how to run them is not.

A campaign that only lives on one angry Twitter account is not a campaign. A campaign that is coherent across three or more theatres in a short window should be treated as coordinated until proven otherwise.

4. Historical and public context (not a playbook)

Smear is older than the internet. What changed is speed, anonymity, persistence, and the merger of press, search, and AI into one memory system.

Political advertising as public history

The 1964 "Daisy" spot, 1984 "Bear in the Woods," and 1988 Willie Horton coverage are taught because they show how a single framed image can relocate an election from policy to fear. They were overt political communication, not covert commercial dark PR — but they trained generations of operatives in emotional compression.

Post-Soviet black PR

Competitive elections in the 1990s produced a professional class of "political technologists." The toolkit described in the literature includes kompromat, paid articles, and identity-confusion tactics. The important educational point is not the tactic list. It is the normalization of reputation destruction as a service industry, which later exported language and posture into business.

Corporate scandals that were not smears

Volkswagen emissions, Samsung Note 7, Tesla Cybertruck demo, KFC UK chicken shortage: these are usually negative publicity born of real events. Teaching them as "smear campaigns" is sloppy. The response playbook (own the fact, fix the operation, stop arguing with reality) is different from the response playbook for a fabricated attack.

Corporate and celebrity information wars

Public litigation in the 2020s has made alleged "social media campaign" planning documents exhibit material. Whatever any court ultimately finds in a given case, the category is now familiar to judges: messaging calendars, scenario planning, and attempts to shape two opposing online narratives at once. That is the modern evidentiary shape of an alleged campaign. Agencies should assume their own planning documents will one day be read aloud.

Platform-era coordination

Research on coordinated inauthentic behavior (CIB) and coordinated social media manipulation (CSMM) documents a recurring pattern: clusters of accounts that do not look fake one-by-one, but move together in time, language, and targeting. Detection lives at the aggregate layer — timing, reuse, graph structure — not at the single-profile layer.

AI-era shift (2024–2026)

Three things changed the cost curve:

• Generative text makes "a lot of plausible articles" cheap.
• Synthetic audio/video lowers the cost of apparent evidence.
• Answer engines summarize the existing web. A smear that wins the corpus can win the model.

The educational implication: defense is no longer "reply to the journalist." Defense is "what will a model and a search index still say in 90 days."

5. How a campaign looks from the outside

This section is written for analysts. It describes observable phases. It does not describe how to cause them.

Phase A — Reconnaissance residue

Before anything public happens, targets often notice odd precursors: unusual records requests, sudden interest in old litigation, staff being approached, domain lookalikes being registered, or a spike in people-search queries. Alone, none of these prove a campaign. Together, they are a watch condition.

Phase B — Narrative seed

A first artifact appears: a blog post, a "concerned" thread, a complaint, a selective leak, a lawsuit paragraph written for quotation. The seed is usually quotable in one sentence. That sentence is the payload. Everything after it is distribution.

Phase C — Legibility

The seed is rewritten into formats other systems accept: a news-shaped item, a Wikipedia-talk talking point, a review, a petition, a hashtag. Each format is a passport into a different index.

Phase D — Synchronization

Multiple surfaces move inside the same 24–72 hours. Language overlaps more than chance would allow. New accounts or rarely used accounts become productive. Journalists receive the same "tip" from more than one direction.

Phase E — Institutionalization

A second-order source cites the first-order source. Google News, aggregators, and later AI systems treat the repetition as confirmation. This is the lock-in phase. After this, "there is no story" is no longer an available line.

Phase F — Maintenance

Drips, "updates," anniversary posts, SEO refresh of the original artifact, replies that keep the thread alive. The goal is not novelty. The goal is to prevent the scar from falling off page one.

Analysts should timestamp every artifact and draw a simple graph: who cited whom. A tree that all roots in one anonymous post is a different object from a tree that roots in a court filing and a named newspaper of record.

6. Surfaces and "tools" — what defenders may monitor, not what attackers should use

You asked for tools. The only responsible list is a defender and researcher stack. An attacker tool list will not be provided.

6.1 Monitoring and intelligence (defense)

• Media intelligence: Meltwater, Cision, Signal AI, Talkwalker, Brandwatch, and comparable suites for print, broadcast, and digital.
• Social listening: native platform search plus enterprise listening; CrowdTangle-class firehose access where still available to researchers; academic tools for coordination detection.
• Search monitoring: Google Search Console (owned properties), rank trackers for branded and attack-adjacent queries, alerts on name + scandal tokens.
• Review and complaint monitoring: Trustpilot, Google Business, industry boards, BBB-type bodies, app stores — as early-warning, not as a place to game.
• Wikipedia and Wikidata watchlists for biographical and company articles.
• Domain and brand abuse: registrar alerts, lookalike domain watch, takedown workflows through counsel.
• AI answer monitoring: weekly protocol — ask major assistants the same ten questions about the client and archive the answers. Treat drift as an incident.
• Legal docket alerts: PACER / local equivalents, regulatory comment dockets.

6.2 Verification (defense)

• Reverse-image and media forensics at a high level (is this the original file, the original date, the original context).
• Corporate and court-record primary sources rather than screenshots.
• Account-history review: creation date, posting cadence, network overlap. One fake-looking account is not proof. A cluster with shared timing is a lead.
• Chain-of-citation mapping: every repetition that looks independent but quotes the same unsourced clause.

6.3 What will not be listed

Software and services whose primary purpose is inauthentic amplification, identity fabrication, non-consensual synthetic media of real people, anonymous mass publishing of injurious claims, or concealment of a campaign sponsor. If a vendor markets "the target will never know who paid," that is not a PR tool. That is a liability.

7. Law: the lines that actually exist

This is not legal advice. Jurisdictions differ. Counsel must localize. The map below is what agencies need in their heads before they opine.

7.1 Defamation

Core idea across common-law systems: a published false statement of fact that injures reputation. Opinions are more protected than facts. Truth is generally a defense. Public figures often face a higher bar (actual malice in U.S. New York Times v. Sullivan doctrine). Private figures usually do not.

Educational consequences:

• "People say X is a predator" can still be treated as a fact assertion depending on wording and context.
• Repeating a lie is often publishing a lie.
• "Anonymous blog, so no defendant" is frequently false. Courts can order platforms and ISPs to identify users. Unmasking actions are now a standard path in online defamation practice.

7.2 Impersonation and synthetic identity

Some U.S. states criminalize digital impersonation of a real person with intent to harm (e.g., California Penal Code § 528.5). Adjacent statutes cover harassing electronic communications and extortionate threats. Deepfake-specific rules are expanding, especially around elections and non-consensual sexual imagery. Do not assume "it's just content."

7.3 Coordinated inauthentic behavior and platform rules

Major platforms prohibit coordinated deception about identity and origin. Enforcement is uneven. A platform ban is not a legal victory, but it is evidence and it can cut distribution. Researchers criticize the vagueness of "inauthentic," and argue the real wrong is manipulative intent — deceiving the audience about who is speaking and why.

7.4 Securities, competition, and consumer law

If the target is a listed company, false campaigns can implicate market-manipulation rules. If the sponsor is a competitor, unfair-competition and business-tort theories appear. Fake reviews are a separate enforcement area in multiple jurisdictions.

7.5 Privacy, doxxing, and harassment

Publishing private addresses, family details, or intimate data to intimidate is not "PR." It is a different legal object and, in many places, a crime. This brief will not discuss how such material is gathered or spread.

7.6 SLAPP and anti-SLAPP

The other side of the board: powerful actors sometimes use lawsuits to silence legitimate criticism. The EU Anti-SLAPP Directive (2024/1069) and various U.S. state anti-SLAPP statutes exist because both smears and abusive countersuits distort the public sphere. A serious agency holds both problems at once. Defense against a smear is not a license to threaten journalists who are doing real work.

7.7 "Verified defamatory information" is not a category

Defamation is injury by false factual statement. If the information is true and lawfully obtained, calling it "defamatory" is a marketing contradiction. If it is false, calling it "verified" is a second problem. A public-facing agency that advertises the production of defamatory content is advertising a tort.

8. How to tell criticism from a campaign

Agencies lose clients by calling every bad story a smear, and they lose clients by missing an actual one. Use a scoring sheet, not a vibe.

Signals that lean toward legitimate negative publicity

• Named, on-the-record sources with first-hand knowledge
• Primary documents that check out
• The outlet has a correction record and an editor
• The target's own conduct created the facts
• Coverage is proportional to the evidence
• Critics disagree with each other (organic controversy is noisy)

Signals that lean toward coordination

• Same uncommon phrase appearing in many "unrelated" places within hours
• A sudden graph of new or dormant accounts
• Tips arriving at several desks with identical exhibits
• Facts that collapse on first contact with a primary source, yet keep circulating
• Mirror sites and syndication that exist only to occupy search real estate
• Timing locked to a deal window
• Refusal of every named person to go on the record, combined with certainty of tone

No single signal is dispositive. Three or more, plus a deal window, is enough to stand up a war room.

9. Effects — what the literature and practice agree on

• Search fossilization. Negative items can reach page one in days and still occupy real estate years later. Time helps, then stops helping.
• Displacement. Positive assets built over years can fall off page one in a week.
• AI inheritance. If the web's high-authority slice is contaminated, assistants will brief third parties with the attack frame.
• Affective polarization. Dirty and dark tactics deepen distrust, including toward the sponsor's own camp over time. Negative campaigning research is mixed on electoral payoff; dirty campaigning's civic cost is clearer.
• Secondary economic harm. Hiring, banking, insurance, partnerships, and inbound press all reprice the target after page one changes.
• Internal trauma. Staff, family, and boards experience smear as persecution. That produces bad official statements if no protocol exists.

Disinformation research commonly finds false or emotional material travels faster than correction. Correction is necessary and usually insufficient by itself.

10. A responsible agency protocol

This is the part that belongs on a consultancy site.

10.1 Before anything happens

• Inventory the six reputation assets in Section 2.
• Pre-clear a holding statement that does not repeat the smear's verb.
• Know counsel, insurer, platform contacts, and a forensic linguist/OSINT analyst before 2 a.m.
• Decide in writing what you will never do in response (fake accounts, fake reviews, counter-smears, paying for retaliation coverage). Put it in the client contract.
• Run tabletop exercises: true scandal, mixed scandal, pure fabrication, leaked docket, synthetic audio.

10.2 First 72 hours

• Preserve. Screenshots are not enough. Hash pages, save HTML, note URLs, times, and who sent what. Litigation holds if counsel says so.
• Triage truth. Three buckets only: true / mixed / false. If you do not know, you do not tweet.
• Stop the self-inflicted wound. No joke replies, no "see you in court" theatre, no repeating the allegation in the denial.
• One spine, many cuts. A single factual spine. Different lengths for staff, investors, press, social, regulators.
• Do not launder the smear. Linking to it, quoting it in full, or SEO-targeting the smear's exact phrase on an official page can help it rank. Answer the underlying question in clean language.

10.3 Days 4–30

• Corrections and retractions where facts are wrong.
• Quiet briefings to reporters who were used, with documents, not vibes.
• Platform reports where rules were actually broken (impersonation, inauthentic coordination, non-consensual synthetic media).
• Search and AI monitoring with a written log.
• If the facts are bad: fix the operation. No communications strategy outruns an unfixed fact.

10.4 30–180 days

• Build durable, independent, accurate pages that deserve to rank for the client's real name and work, not for the smear's vocabulary.
• Recover relationships the whisper network damaged.
• After-action: what precursor was missed.

10.5 What "winning" means

Not "the internet forgot." Winning is: page one is no longer only the attack; AI answers include the correction; counterparties will still take the meeting; staff will still stay; the next journalist starts from a fuller file.

11. Ethics for the profession

Public relations, as taught by its own codes (PRSA, CIPR, Global Alliance), is the management of relationships through disclosed, accurate enough to be honorable communication. A practice whose product is hidden harm to a third party is not a specialty of PR. It is an adjacent underground.

Teaching smear as a service offering does three things:

• It invites civil defendants and, in some fact patterns, prosecutors.
• It makes every legitimate client toxic by association.
• It trains juniors that concealment is a craft rather than a red flag.

The professionally useful expertise is the opposite one: know the anatomy well enough to recognize it, document it, brief counsel, brief a newsroom, and keep a client from answering a smear with a smear.