The short version
- In the United States, platforms are generally not treated as the publisher of what users post (47 U.S.C. § 230). The author can be liable; the platform usually cannot.
- A court can order an author to remove a defamatory post, but courts have held that Section 230 can stop them ordering the platform to do so (Hassell v. Bird).
- Anonymous authors are not simply unmasked. Courts require notice and a substantial evidentiary showing first (Dendrite, Cahill).
- Opinion and statements that cannot be proven false are protected; provable false statements of fact are not (Milkovich).
- Public figures face a higher bar, including proving actual malice (New York Times v. Sullivan, Gertz).
- Copyright is separate: the DMCA notice-and-takedown system works for works you own, and misuse carries liability (17 U.S.C. § 512, Lenz).
- In the EU and UK, privacy law can require search engines to delist results (Google Spain, GDPR Article 17), balanced against free expression.
How to use this page
Each entry gives the court or body, the year and the citation, a plain summary, why it matters for removal, and links to the primary or authoritative source. This is general information, not legal advice. Laws differ by country and change over time, so check the current position or speak to a qualified lawyer before acting.
Who is legally responsible: the author or the platform?
The starting point in the United States is Section 230 of the Communications Decency Act (47 U.S.C. § 230). It provides that a provider or user of an interactive computer service is not to be treated as the publisher or speaker of information provided by someone else. In practice, the person who wrote a defamatory comment can be sued. The platform that hosts it generally cannot.
The leading early decision is Zeran v. America Online (4th Cir. 1997). Kenneth Zeran had his phone number attached to offensive T-shirt ads in anonymous posts. He told AOL, but the posts stayed up, and the court held that Section 230 barred liability even after notice. That is why the idea of "tell the site, then sue if they refuse" is not a reliable strategy against a US platform.
Hassell v. Bird (California Supreme Court, 2018) went a step further. A law firm won a default judgment over defamatory reviews and an order directing Yelp to remove them. The court held, in a 4 to 3 decision, that Section 230 stopped the court from ordering Yelp to remove the reviews. The judgment against the reviewer stood, but the platform was not compelled.
Does v. Reddit (9th Cir. 2022) applied the same principle to Reddit itself, in a case about child sexual abuse material posted by users. The Ninth Circuit held that Section 230 protected Reddit and read the statute's sex-trafficking exception narrowly. The U.S. Supreme Court declined to review it in 2023. The practical result is that legal pressure on Reddit over what users wrote rarely works, and removal is normally pursued through Reddit's own rules and reporting channels.
What defamation law actually covers
Defamation law varies by country and, in the United States, by state. In broad terms, a claim involves a false statement of fact about a person or business, communicated to others, made with the required level of fault, and causing harm. Truth is a defence, and so are some categories of opinion.
Milkovich v. Lorain Journal (U.S. Supreme Court, 1990) is the key case on opinion. The Court declined to create a blanket privilege for opinion, but confirmed that statements which cannot reasonably be read as stating actual facts, or which cannot be proven false, are protected. "This company is terrible" is opinion. "This company kept my deposit" is a factual claim that can be true or false. Only the second can be challenged.
Public figures face a higher bar
New York Times v. Sullivan (1964) requires public officials to prove actual malice, meaning the statement was made knowing it was false or with reckless disregard for the truth. Gertz v. Robert Welch (1974) lets states allow private individuals to recover on a showing of fault, while limiting presumed and punitive damages unless actual malice is shown, and it draws the line between public figures and private people. Founders and executives of well-known companies are sometimes treated as public figures for related statements, which raises the bar for a claim.
Anonymous authors
Many negative posts are written under usernames. Courts do not hand over an author's identity simply because someone asks. In Dendrite International v. Doe No. 3 (New Jersey, 2001) and Doe v. Cahill (Delaware, 2005), courts set the modern tests for unmasking an anonymous speaker. The plaintiff must notify the speaker, identify the exact statements complained of, and produce real evidence supporting the claim before a court will order an identity disclosed.
For a business or individual, this has two consequences. First, unmasking is a legal process that needs a lawyer, not a report to a platform. Second, it is rarely quick or cheap, and it only makes sense when the statements are clearly false statements of fact and the harm is serious. We do not try to identify anonymous users.
Copyright and the DMCA
Copyright is a separate legal route. Under 17 U.S.C. § 512(c), a platform that receives a valid notice and acts promptly on it keeps its safe harbour. A valid notice must identify the copyrighted work and the infringing material, give the sender's contact details, include a good-faith statement, and include a statement under penalty of perjury that the notice is accurate and the sender is authorised.
Section 512(f) makes knowing material misrepresentation in a notice actionable. In Lenz v. Universal Music (9th Cir. 2015), the court held that copyright owners must consider fair use before sending a takedown notice, and that a notice sent without that consideration can support a misrepresentation claim. A DMCA notice is therefore a legal statement and should be checked before it is sent.
Copyright protects original works such as photographs, writing and video. It does not protect facts, and it cannot be used to remove criticism or reviews that do not copy your work.
Privacy and the right to be forgotten
Privacy law works differently. In Google Spain v. AEPD and González (Court of Justice of the EU, 2014), the Court held that a search engine operator is responsible for the personal data in the results it lists. Individuals can ask for links to be removed when the data is inadequate, irrelevant, no longer relevant or excessive, unless there is a public interest, such as the person's role in public life.
Article 17 of the GDPR, and its UK equivalent, gives individuals a right to ask for personal data to be erased in set circumstances, for example when it is no longer necessary or was unlawfully processed. It does not apply where processing is necessary for freedom of expression and information, for legal obligations or for legal claims. The right acts on search listings and on data controllers, not on the underlying Reddit post, and it does not turn unwelcome commentary into removable content.
Google also has its own policy on personal contact information and doxxing. It lets people request removal from Search of certain personal details such as addresses, phone numbers, email addresses, government ID numbers and bank or card numbers, and of doxxing content that combines contact details with threats or calls to harass. Removing a Reddit post fixes the source, and Google's tools can speed up removing the same details from search results.
Reviews, consumer law and market rumours
Some regulators' guidance is relevant because it explains when criticism is legitimate. The FTC's Rule on the Use of Consumer Reviews and Testimonials (16 CFR Part 465), effective October 21, 2024, prohibits fake or false consumer reviews and testimonials, buying or conditioning reviews, undisclosed insider reviews, and review suppression through threats or intimidation. It cuts both ways: a business cannot seed fake praise, and it should not threaten reviewers to silence them. Honest evidence that a review is fake is a legitimate basis for a report.
The FTC's staff guidance on multi-level marketing explains why scrutiny of income claims is a recognised consumer-protection topic. Criticism of those claims is commentary on a regulated area, not content that is removable just because it is unwelcome. The SEC's investor alert on social media and stock rumours warns that fraudsters can spread false rumours to move share prices. Coordinated false rumours about a listed company are a matter for the SEC and counsel, and a Reddit report cannot resolve market manipulation.
Court orders and Reddit outside the US
Laws differ by country, and Reddit's transparency reports describe requests from courts and authorities around the world. Reddit reviews each request against its own rules first and may restrict content in one country rather than deleting it everywhere. Its reports describe, for example, court orders that led to content being restricted in a single country while Reddit disagreed with the order and appealed.
This matters because the outcome of a legal route is not always global deletion. Depending on the legal basis and the jurisdiction, the result can be global removal, geographic restriction, removal of individual comments, removal of a post, community moderation, or no action at all.
What this means in practice
- Start with Reddit's rules. They are the fastest and most reliable route for content that breaks them.
- Use copyright only for works you own, and check the notice carefully.
- Use privacy routes for personal data, and expect different results from search engines and from Reddit.
- Treat defamation as a legal question for a lawyer, especially where the author is anonymous or the subject is a public figure.
- Keep records. Dated screenshots, permalinks and every report and reply matter in every route.