Defamation and Content Liability Toolkit: Review, Retraction, Anti-SLAPP, and Coverage

By Casey Scott McKay · 2024-05-13

Copyright tells a publisher what it may reproduce. It says nothing about the far larger risk of being sued for what a piece actually says. This toolkit assembles the working material for that risk: the elements of a defamation claim and where each is won or lost, the fault standards that turn on the plaintiff's status, the privileges and defences that protect reporting, and the pre-publication review process that produces the record on which everything later depends. It covers the retraction statutes that limit damages, the anti-SLAPP procedures that dispose of weak claims early and vary enormously by jurisdiction, the insurance that responds and the notice conditions that forfeit it, and the adjacent torts — false light, trade libel, and privacy claims — that arrive in the same complaint.

IP and Technology > Internet | Toolkit | Published 13 May 2024 - Updated 19 June 2026 | Casey Scott McKay - marksy.us

Summary. Copyright tells a publisher what it may reproduce. It says nothing about the larger risk of being sued for what a piece says. This toolkit covers the elements of a defamation claim and where each is won or lost, the fault standards that turn on the plaintiff's status, the privileges protecting reporting, the pre-publication review that produces the record everything later depends on, the retraction statutes that limit damages, the anti-SLAPP procedures that vary enormously by jurisdiction, the insurance that responds and the notice conditions that forfeit it, and the adjacent torts arriving in the same complaint.

Keywords: defamation · publisher liability · pre-publication review · actual malice · public figure · retraction statutes · anti-SLAPP · media insurance · false light · trade libel · product disparagement · source protection · republication · single publication rule · substantiation files

Start Here

Anyone who publishes carries two categories of legal risk. The first is intellectual property: whether the material reproduced belongs to someone else. The second is content liability: whether what is said causes actionable harm to a person or a business. The first is well served by the rest of this collection. The second is larger, is faced by every publisher including those who never reproduce anything, and is the subject of this toolkit.

The core claim is defamation: a false statement of fact, of and concerning the plaintiff, published to a third party, causing reputational harm, with a fault standard that depends on who the plaintiff is.

Each element is a place to win.

Falsity. Truth is a complete defence, which makes the substantiation file the single most valuable document a publisher creates. Substantial truth suffices; minor inaccuracies that do not change the sting of the statement do not create liability.

Fact rather than opinion. Statements incapable of being proven true or false, and statements that a reasonable reader would understand as commentary, are not actionable — but labelling something an opinion does not make it one, and an opinion implying undisclosed defamatory facts remains actionable.

Of and concerning. The statement must be understood to refer to the plaintiff. Group references, composites, and unnamed subjects raise identification questions that are frequently the whole case.

Publication. Communication to a third party, with republication rules and the single publication rule governing when the clock starts and how many claims exist.

Fault. Public officials and public figures must prove actual malice — knowledge of falsity or reckless disregard for the truth — under New York Times Co. v. Sullivan and its progeny. Private figures on matters of public concern must generally show at least negligence, per Gertz v. Robert Welch, Inc..

Damages. Presumed damages are constrained where the actual malice standard is not met, per Gertz and Dun & Bradstreet, Inc. v. Greenmoss Builders, Inc., and special damages must be pleaded for some categories.

See The Sentence That Costs You for the doctrinal treatment and Managing Defamation and Content Liability for the operational sequence.

Who the plaintiff is, and why it decides the case

The fault standard is the most outcome-determinative variable in defamation practice, and it turns on a classification made at the outset.

Public officials are those with substantial responsibility for governmental affairs. The actual malice standard applies to statements about their official conduct.

All-purpose public figures are people of such pervasive fame or notoriety that they are public figures for all purposes. The category is narrow.

Limited-purpose public figures have thrust themselves into a particular public controversy to influence its outcome, and the standard applies to statements relevant to that controversy. This is where most contested classifications sit, and the analysis asks whether there was a public controversy, whether the plaintiff voluntarily entered it, and whether the statement relates to their participation.

Involuntary public figures are theoretically possible and practically rare.

Private figures need show only the fault standard the jurisdiction requires, commonly negligence, on matters of public concern — and possibly less on matters that are not.

Corporations and businesses can be defamed, and are frequently limited-purpose public figures with respect to controversies about their products or conduct.

The practical instruction for a publisher is to make the classification during review and to record it, because the level of verification appropriate to a story is a function of the standard that will apply to it. A publisher that assumed a subject was a public figure and was wrong has applied the wrong verification standard to the wrong story.

Privileges, defences, and the reporting protections

Beyond truth and opinion, a set of privileges protects specific categories of publication, and knowing which applies is often the fastest route to disposal.

The fair report privilege protects fair and accurate reports of official proceedings and records. Its scope varies by jurisdiction, its accuracy requirement is real, and it can be defeated by embellishment beyond what the record says.

Neutral reportage protects accurate reporting of accusations made by a responsible party about a public figure, and is recognised in some jurisdictions and not others.

The wire service defence protects republication of material from a reputable source absent notice of falsity, and is similarly jurisdiction-dependent.

Absolute privileges protect statements in judicial and legislative proceedings, which matters for what may safely be repeated from a filing.

Qualified privileges protect communications made in the discharge of a duty or in a common interest, and are defeated by malice.

Consent bars a claim by a subject who agreed to publication.

The incremental harm and libel-proof plaintiff doctrines limit recovery where the challenged statement adds nothing to reputational damage already caused by true statements.

Statutory intermediary protections shield platforms from liability for content supplied by others, which is treated in the Platform Liability and Section 230 Toolkit and matters here because a publisher that hosts comments occupies both roles.

The single publication rule treats a mass communication as one publication for limitation purposes, with republication reviving the claim — which makes decisions about updating, reposting, and re-promoting archived material legally significant.

Pre-publication review, which is where cases are won

Litigation about a published piece is decided largely by decisions made before publication, and the review process is the mechanism that makes those decisions defensible.

Review the statements, not the article. Extract each factual assertion about an identifiable person or business and treat it individually.

Source each assertion. Named source, document, record, observation, or inference. Assertions supported only by inference need to be presented as inference.

Check the level of verification against the standard. A statement about a private figure requires more care than one about a public official, because negligence is easier to prove than actual malice.

Attend to the sting. Liability follows the defamatory meaning a reasonable reader takes, not the literal words. A technically accurate assembly of facts implying something false is actionable.

Watch the headline, the caption, and the social copy. These are read by more people than the article, are written by different people, and routinely overstate what the piece supports.

Seek comment. Putting allegations to the subject before publication is both good practice and evidence against reckless disregard, and the response should be recorded whether or not it is used.

Record the decisions. Who reviewed, what was changed, what was verified, and what was consciously left in. This file is the defence.

Do not create documents that read badly. Internal messages speculating about whether something is true, or joking about the subject, are discoverable and are the single most damaging category of evidence in defamation litigation.

Handle images and video separately, since a photograph juxtaposed with text can carry a defamatory implication the text does not.

Preserve the file in accordance with a retention policy that survives a litigation hold under FRCP 37.

Retraction, correction, and the demand letter

Most content disputes never reach a filing, and the response to a complaint determines whether they do.

Retraction statutes exist in many jurisdictions and typically limit a plaintiff's recoverable damages where a timely and adequate correction is published following a proper demand. The conditions are technical: the demand must meet statutory requirements, and the correction must be published within a defined period, with defined prominence.

A correction is not an admission where handled properly, and the commercial and reputational value of correcting a genuine error usually exceeds the litigation value of defending it.

Speed matters. Statutory windows are short and the practical window is shorter, since a subject who receives no response escalates.

Distinguish the categories. A factual error, a disputed characterisation, a complaint about tone, and a demand to remove accurate but unwelcome information are four different problems with four different responses.

Removal is not the default. Depublishing accurate reporting on demand creates an expectation and, for a news organisation, a professional problem. Correction, clarification, and updating are usually better.

Archives and updates raise republication questions, since substantive alteration may restart limitation.

Preserve everything on receipt of a demand, since the hold obligation attaches earlier than most editorial teams assume.

Involve the insurer early, because policies commonly require notice of circumstances that may give rise to a claim, and a settlement or correction made without notice can prejudice cover.

Anti-SLAPP, and why the forum decides the cost

Anti-SLAPP statutes permit early dismissal of claims arising from protected speech, typically with fee shifting, and their availability is the largest single determinant of what a defamation defence costs.

The structure is usually two-step. The defendant shows the claim arises from protected activity; the plaintiff must then show a probability of prevailing, on evidence rather than pleading.

Fee shifting is the point. A successful motion commonly carries mandatory fees, which changes the settlement dynamics of every marginal claim.

Discovery is usually stayed pending the motion, which is where most of the cost savings arise.

Immediate appeal is available in some jurisdictions, which cuts both ways.

Coverage varies enormously. Some jurisdictions have broad statutes reaching any speech on a matter of public interest; some have narrow ones limited to petitioning activity; some have none.

Application in federal court is contested, with courts differing on whether these procedures apply in diversity cases, which makes forum a live strategic question.

Plaintiffs forum shop accordingly, and defendants should consider removal, transfer, and dismissal sequencing with the anti-SLAPP position in mind.

Timing is strict. Motions must generally be filed within a short window after service, and missing it forfeits the mechanism.

See the Content Liability Checklist for the assessment sequence.

The adjacent torts that arrive in the same complaint

A defamation complaint rarely contains only defamation, and the additional counts have different elements, different defences, and different insurance consequences.

False light protects against highly offensive false implications, is recognised in some jurisdictions and rejected in others, and can survive where a defamation count fails because the statements were literally true.

Trade libel and product disparagement concern statements about a business's goods or services rather than its reputation, generally require proof of special damages, and overlap with false advertising claims under 15 U.S.C. § 1125 where the parties are competitors.

Intrusion and public disclosure of private facts are privacy torts reaching truthful publication, with a newsworthiness defence whose scope is contested.

Appropriation and right of publicity claims arise where a person's name or likeness is used commercially. See Your Face Is Not Public Domain and the Right of Publicity and Personal Brand Toolkit.

Intentional infliction of emotional distress is pleaded alongside and is constrained where the speech is protected, per Hustler Magazine, Inc. v. Falwell.

Tortious interference claims arise where a publication causes a contract or relationship to fail.

Negligent misstatement and product-review liability arise where a publisher's factual assertions are relied on to a reader's detriment.

Copyright and trademark counts appear where the piece reproduced material or used a mark, which is where this toolkit meets the rest of the collection. See the Copyright Enforcement Toolkit and The First Amendment and Trademark Toolkit.

Insurance, and the conditions that forfeit it

Media liability insurance is the mechanism by which most publishers survive a serious claim, and the conditions attached to it are routinely breached before anyone reads the policy.

Notice of circumstances is required by most policies, and a demand letter is a circumstance. Handling it internally for three months and then tendering is the standard way cover is lost.

Consent to settle provisions mean a publisher that corrects, apologises, or pays without insurer consent may prejudice indemnity — which conflicts directly with the commercial instinct to resolve quickly.

Panel counsel requirements restrict who may act, and a publisher with a preferred adviser should negotiate this at placement rather than at claim.

Prior acts and retroactive dates determine whether an older publication is covered, which matters because archives generate claims years after publication.

Defence costs inside or outside the limit is the single most consequential coverage term, since defence costs in this area frequently exceed any damages.

Punitive damages may be uninsurable in some jurisdictions.

Intentional acts exclusions interact awkwardly with the actual malice standard, and a finding of knowing falsity may fall outside cover.

Assumed liability exclusions matter where a publisher has indemnified a distributor or platform.

See the IP Insurance and Risk Transfer Toolkit.

Contributors, freelancers, and who carries the risk

Publishers rarely write everything themselves, and the allocation of content liability across contributors is frequently mismatched to the parties' ability to bear it.

Freelance agreements typically contain warranties and indemnities running from the writer to the publisher, covering accuracy, originality, and non-infringement. Those indemnities are worth the freelancer's assets, which is usually nothing.

Insurance rarely follows the indemnity. A freelancer will not have media liability cover; the publisher will. Requiring an indemnity the contributor cannot honour is theatre, and the useful term is instead a warranty of source materials and cooperation.

Editorial control determines exposure. A publisher that commissions, edits, and headlines a piece is the publisher of it, whatever the contract says between the parties.

User-generated content sits differently, under the intermediary protections, provided the publisher has not materially contributed to the unlawful content.

Syndication and republication spread the exposure, and each republisher is generally liable for its own publication.

Contributor conduct — undisclosed conflicts, fabrication, plagiarism — is a professional risk with legal consequences, and verification processes should not assume good faith merely because a contributor is known.

Anonymous and confidential sources raise the question of what can be proved at trial if the source will not testify. A story that cannot be defended without burning a source is a story with a hidden cost.

Source protection varies by jurisdiction and by court, and a publisher relying on a shield law should know its scope before it relies on it rather than after.

Publishing across borders

Defamation is one of the least harmonised areas of law, and cross-border publication multiplies exposure rather than averaging it.

Some jurisdictions place the burden of proving truth on the defendant, which inverts the position and makes the substantiation file even more important.

Fault standards differ. The actual malice requirement is not universal, and a public figure elsewhere may face a much lower bar.

Damages differ, and jurisdictions with lower awards may nonetheless produce injunctive relief that is commercially worse.

Jurisdiction attaches to accessibility in several regimes, meaning a piece published anywhere and readable somewhere may be sued upon there.

Enforcement of foreign judgments is constrained in some countries by statutes that decline recognition where the foreign law offers less speech protection.

Prior restraint availability varies, and injunctions preventing publication are obtainable in some places and effectively unavailable in others.

Corporate structure matters. Which entity publishes, where it is established, and where the servers sit all affect the forum analysis.

Practical management consists of a jurisdiction map for the publication's main markets, a review standard set to the strictest of them for material likely to be read there, and legal contacts in each.

See the Cross-Border IP Litigation Toolkit.

Investigative work, and the risk that is worth taking

Everything in this toolkit could be read as an argument for publishing less, which would be a misreading. The point of a content liability practice is to make difficult publication possible, not to prevent it.

Investigative reporting is where the risk concentrates and where the value is. Six practical points distinguish work that survives a claim from work that does not.

Build the file as you go. Reporters who assemble documentation contemporaneously produce a defence for free. Reporters who reconstruct it after a demand produce something that looks reconstructed.

Lock the sting early. Decide what the piece actually asserts, in a sentence, before it is written. Most defamation exposure comes from implications the writer did not intend and the editor did not notice, and naming the assertion at the outset makes it reviewable.

Verify to the standard that will apply, not to the standard that feels sufficient. A private individual caught up in a story about an institution is a private figure, and the negligence standard is unforgiving.

Give the subject a real opportunity to respond. Not a formality on deadline. A substantive letter, with adequate time, putting each allegation. This is both fairer and the strongest available evidence against reckless disregard, and it frequently improves the story.

Do not overreach in the promotion. Careful copy in the piece and an unsupportable headline is the most common structural failure, and the headline is what a court will read first.

Decide the source position before publication. If the story cannot be defended without testimony from a source who will not give it, that is a decision to be taken deliberately by someone senior — not a discovery made two years later by a litigator.

Publishers that do those six things publish difficult stories and win the cases that follow. Publishers that treat legal review as a gate to be passed rather than a process to be used publish the same stories with none of the protection, and settle claims that were defensible.

The corollary is worth stating plainly to editorial leadership: the purpose of the review process is not to reduce the number of hard stories. It is to make sure that when one of them is challenged, the file supports it — which is precisely what allows the next hard story to be commissioned.

The non-media publisher

Most content liability advice is written for news organisations, and most people who need it are not one. A company blog, a research firm's reports, a trade association's newsletter, an analyst's notes, a reviewer's site, and an executive's social account all publish, and none of them has a legal desk.

The differences that matter are five.

No review process exists. Publication happens on a marketing calendar, approved by people who have never considered defamation. The single highest-value intervention is a one-page rule: any factual assertion about a named competitor, customer, supplier, or individual goes to legal before it publishes. Nothing else needs to change.

The speech may be commercial. A company disparaging a competitor's product is exposed to trade libel and to false advertising under 15 U.S.C. § 1125, and the constitutional protections that shield reporting apply differently to commercial speech.

There is usually no media insurance. General liability policies contain personal and advertising injury cover that may respond to some publication claims, and its scope is narrower than a media policy. Check it before assuming there is cover.

Employees publish. An executive's post, a salesperson's message, and a support agent's comment are all publications by someone whose statements may be attributed to the company. Social media policy is content liability policy.

Reviews and ratings are the recurring flashpoint. A business that publishes comparative assessments of named products invites both trade libel claims and, from the other direction, pressure to remove accurate criticism. See the Advertising and Marketing Law Toolkit.

For a non-media publisher, the whole of this toolkit reduces to four things: the one-page rule about naming others, a substantiation habit for any factual claim, a check on what the general liability policy actually covers, and a social media policy that treats employee publication as company publication. That is an afternoon of work and it addresses the great majority of the exposure.

A short glossary

Defamatory meaning. The sense a reasonable reader takes from the publication, which may differ from its literal words and is what liability follows.

The sting. The essential defamatory charge. Substantial truth of the sting defeats the claim even where details are wrong.

Substantial truth. The doctrine that minor inaccuracies not changing the sting do not create liability.

Of and concerning. The requirement that the statement be understood as referring to the plaintiff. Frequently the whole dispute where the subject is unnamed.

Actual malice. Knowledge of falsity or reckless disregard for the truth. A subjective standard about the publisher's state of mind, not a measure of ill will.

Limited-purpose public figure. Someone who entered a public controversy to influence it, and who therefore faces the actual malice standard for statements about that controversy.

Single publication rule. One mass communication is one publication for limitation purposes; republication starts a new clock.

Republication. Substantive re-issuance of material, which can revive a claim — a live question for archives and updates.

Fair report privilege. Protection for fair and accurate accounts of official proceedings and records.

Neutral reportage. Protection, recognised in some jurisdictions, for accurate reporting of accusations by a responsible source.

Retraction statute. A provision limiting recoverable damages where a timely, adequate correction follows a proper demand.

Anti-SLAPP. Procedure permitting early dismissal of claims arising from protected speech, usually with fee shifting and a discovery stay.

False light. A privacy tort covering highly offensive false implications, recognised in some jurisdictions and rejected in others.

Trade libel. Disparagement of goods or services rather than of reputation, generally requiring special damages.

Notice of circumstances. The insurance condition requiring a publisher to report a demand that may become a claim. The most commonly breached term in media policies.

Practitioners who keep those fifteen straight will avoid the field's characteristic errors: treating truth as a complete answer without asking what the sting is, treating an opinion label as protective, and treating a demand letter as a matter to be handled quietly before telling the insurer.

Sequencing a live matter

When a demand or a claim arrives, the order of operations matters more than the analysis, because several of the steps have deadlines and one of them protects everything else.

Hour one: preserve. Issue the hold to editorial, legal, and communications, covering document management systems, email, messaging platforms, and personal devices used for work. Preservation obligations attach on reasonable anticipation of litigation, which a demand letter supplies, and failures here are unforgivable under FRCP 37.

Day one: tender. Notify insurers. Read the policy's notice provision rather than assuming a timeframe, and record the date and method of notice.

Day one: freeze the piece. No edits, no quiet corrections, no removal, until the position is assessed. An unrecorded alteration after a demand is the worst available fact.

Days one to three: triage. Factual error, disputed characterisation, tone complaint, or removal demand. Only the first calls for a correction on the merits.

Days one to five: assemble the file. The statement extraction sheet, the sources, the right-of-reply record, and the review notes. If these do not exist, the matter's difficulty has just doubled.

Week one: classify the plaintiff and check limitation. Public figure status and the single publication rule frequently dispose of a claim before any merits analysis.

Week one: assess the retraction statute. If a correction is warranted, statutory windows are short and the damages limitation is worth having.

Week two: assess the forum and the anti-SLAPP position before responding substantively, because the response may affect where the claim is brought and the motion deadline is short.

Throughout: control the internal conversation. Everything written about the matter is discoverable. Legal advice is privileged; speculation in a group chat is not.

A publisher that runs this sequence competently converts most demands into corrections or into nothing. A publisher that runs it badly turns a defensible piece into a settlement, usually because of something done in the first week.

A closing observation

Defamation is the area of law where practitioners most often give advice that is technically correct and practically useless, because the answer to almost every question is "it depends on the record," and the record was made months earlier by people who were not thinking about litigation.

That has a consequence for how this work should be organised. The valuable hours are not the ones spent analysing a demand letter; they are the ones spent, long before, designing a review process that produces a defensible file as a by-product of ordinary editing. A statement extraction sheet costs a reporter twenty minutes and is worth more in a dispute than any brief.

The same applies to the two failures that most reliably convert defensible pieces into settlements: the headline nobody reviewed, and the internal message nobody thought about. Both are addressed by process rather than by doctrine, both are cheap to fix, and neither is anybody's job unless someone makes it so.

And there is a broader point worth making to any publisher, media or otherwise. Content liability is asymmetric: the cost of defending a weak claim can exceed the cost of settling it, which means claimants with no case can extract money from publishers with a complete defence. Anti-SLAPP statutes exist to correct that asymmetry and do so unevenly. Insurance corrects it further and is conditional. A good file corrects it most, because it converts a two-year discovery process into an early motion.

Build the file. Everything else in this toolkit is what to do when the file is thin.

One final caution for practitioners advising outside their home jurisdiction. The protections described here are unusually strong by international standards, and a publisher accustomed to them will underestimate its exposure elsewhere. In jurisdictions placing the burden of proving truth on the defendant, with no actual malice requirement and no early-dismissal mechanism, the same piece carries a materially different risk profile — and the practical answer is often not a legal one but an editorial decision about which markets a story is promoted in, taken with the map in front of everyone.

That map — main markets, burden of proof, fault standard, early dismissal availability, and local counsel — fits on a single page and should be produced once for any publication with an international readership. Producing it after a foreign claim arrives is possible; producing it before means the decision that mattered was taken by an editor rather than by a court.

Review it annually, and whenever the publication opens a market or a jurisdiction changes its rules — both of which happen more often than editorial teams expect, and neither of which generates a notification to counsel.

Keep a copy with the insurance schedule, since the two documents answer the same question from different directions: what happens if this piece is challenged, and who pays for finding out.

Neither document is expensive to produce, and a publisher that holds both has already done more than most of its peers to make hard journalism survivable.

Which is, in the end, the whole purpose of a content liability practice: not fewer difficult stories, but difficult stories that hold.

Everything in the preceding pages is in service of that sentence, and a publisher that reads nothing else should read it twice.

Then commission the statement extraction sheet, which is where the reading turns into practice.

A Suggested Reading Path

New to content liability: The Sentence That Costs You, then Managing Defamation and Content Liability, then the Content Liability Checklist.

If you host third-party content: the Platform Liability and Section 230 Toolkit, Deciding What Stays Up, and the Content Moderation and Platform Policy Toolkit.

If the piece uses someone's likeness: the Name, Image, and Likeness Clearance Checklist and the Right of Publicity and Personal Brand Toolkit.

If a competitor's product is discussed: the Advertising and Marketing Law Toolkit and Raising a Trademark Fair Use Defense.

If material is reproduced: Running a Fair Use Analysis, the Fair Use Risk Assessment Checklist, and the Fair Use and Permissions Toolkit.

If the speech is expressive and a mark is involved: The First Amendment and Trademark Toolkit.

Insurance: the IP Insurance and Risk Transfer Toolkit.

Adjacent publishing practice: the Publishing, Photography, and Author Rights Toolkit and the Podcasting and Audio Rights Toolkit.

Primary Authorities

| Authority | Use | |---|---| | New York Times Co. v. Sullivan | Actual malice for public officials | | Curtis Publishing Co. v. Butts | Extension to public figures | | Gertz v. Robert Welch, Inc. | Private figures, fault, and presumed damages | | Dun & Bradstreet v. Greenmoss Builders | Matters not of public concern | | Philadelphia Newspapers, Inc. v. Hepps | Burden of proving falsity | | Milkovich v. Lorain Journal Co. | Opinion implying undisclosed facts | | Hustler Magazine, Inc. v. Falwell | Emotional distress claims and protected speech | | Masson v. New Yorker Magazine, Inc. | Altered quotations and material change of meaning | | Harte-Hanks Communications v. Connaughton | Proving reckless disregard | | St. Amant v. Thompson | Subjective awareness of probable falsity | | Time, Inc. v. Firestone | Limits on limited-purpose public figure status | | Bartnicki v. Vopper | Publication of unlawfully obtained information | | Snyder v. Phelps | Speech on matters of public concern | | Nebraska Press Association v. Stuart | Prior restraint | | Branzburg v. Hayes | Source protection and testimonial obligations | | 15 U.S.C. § 1125 | Trade libel overlapping false advertising | | 15 U.S.C. § 45 | Claims made in commercial publication | | 17 U.S.C. § 107 | Quotation and excerpt in the same piece | | 17 U.S.C. § 512 | Notices concerning hosted material | | 47 U.S.C. § 230 | Intermediary protection for third-party content | | FRCP 12 | Early dismissal alongside anti-SLAPP procedure | | FRCP 26 | Discovery into editorial process | | FRCP 37 | Preservation of the review file | | FRCP 56 | Summary judgment on fault and falsity |

Search the underlying materials directly for actual malice reckless disregard evidence, limited purpose public figure analysis, retraction statute damages limitation, anti-SLAPP federal diversity application, and media liability policy notice condition.

Forms and Templates

A statement extraction sheet, listing each factual assertion about an identifiable person or business, its source, its verification level, and the reviewer's note. This single document is the sector's most valuable and least used instrument.

A subject classification note recording whether each subject is a public official, public figure, limited-purpose public figure, or private figure, and why — because the verification standard follows from it.

A right-of-reply record, capturing what was put to the subject, when, how, and what was said in response, whether or not the response was published.

A headline and social copy review step, separate from the article review, because different people write them and they overstate.

A source protection assessment for confidential sources, addressing what can be proved without them and what the applicable shield protection actually covers.

A demand response protocol with a triage rule distinguishing factual error, disputed characterisation, tone complaint, and removal demand, each with a template first response and a timeline.

A retraction and correction template meeting the statutory requirements of the jurisdictions the publication serves, with prominence and timing rules stated.

An insurance notice trigger, requiring tender on receipt of any demand rather than on filing of any claim, with a named person responsible.

An archive and update policy, addressing what may be altered, what constitutes republication, and what is recorded when a piece is changed.

A contributor agreement with source-materials warranties and cooperation obligations rather than indemnities the contributor cannot honour.

A litigation hold template that reaches editorial systems, messaging platforms, and personal devices used for work.

An internal communications standard, reminding editorial staff that messages about a subject are discoverable — the cheapest risk reduction available and the one nobody implements until after the first case.

For general drafting starting points, see the Draft License Agreement and the License Agreement Template.

Five recurring matters

A demand letter arrives about a story published two years ago. Check limitation and the single publication rule, check whether the piece has been substantively updated in a way that constitutes republication, tender to insurers immediately, and preserve. Then triage the substance: if there is a factual error, correcting it promptly may engage the retraction statute and is usually right on the merits as well.

A subject demands removal of accurate reporting. Removal is not the default and should not become one. Consider updating with subsequent developments, adding context, or noting the subject's position. A publisher that depublishes on demand creates an expectation that every future subject will invoke.

A claim is filed in a jurisdiction with no anti-SLAPP statute. Assess removal, transfer, and forum non conveniens with the fee-shifting position in mind, and be realistic that a marginal claim in an unfavourable forum may be worth settling early precisely because the mechanism to dispose of it cheaply does not exist.

Internal messages surface in discovery. They usually do. The damage is done at the time of writing, not at the time of production, which is why the internal communications standard belongs in onboarding rather than in a crisis.

A freelancer's piece turns out to have been partly fabricated. The publisher is the publisher. Correct comprehensively, notify insurers, review everything else by the same contributor, and expect the indemnity in the freelance agreement to be uncollectable. The lesson is about verification process rather than about contract drafting.

What good looks like

Statements are extracted and sourced individually, not reviewed as prose.

Subject classification is recorded, so the verification standard applied matches the standard that will be applied later.

Comment is sought and recorded as a matter of routine.

Headlines and social copy get their own review.

The review file is preserved and is the kind of document a publisher is content to produce.

Internal messaging discipline exists and is taught, because it is the single most damaging evidence category.

Demands are triaged within days, tendered to insurers on receipt, and answered in a way that engages retraction statutes where a correction is warranted.

The archive policy addresses republication, so that updating an old piece is a considered decision rather than a routine edit.

Publishers with those eight defend claims from a record. Publishers without them defend from recollection, against a plaintiff holding their own internal messages.