What is online defamation?

What counts as online defamation

The same four elements apply as offline: a false statement of fact, published to someone other than you, attributable to a person at fault, causing harm. Those are set out in full on what is defamation of character. Because anything typed into a browser is fixed and recorded, online defamation is almost always libel rather than slander, which is explained on libel vs slander.

Likely a factual assertion	Likely not actionable
"He was arrested for fraud" when he was not	"He seems dishonest to me"
"This clinic is not licensed" when it is	"This clinic was a waste of money"
"They charged my card twice" when they did not	"Worst experience of my life"
"She was fired for stealing" when she resigned	"I would never go back"

The dividing line is not tone. It is whether the statement could be checked and shown to be false.

Internet defamation and the platform question

The first instinct is to write to the website. In the United States that route mostly closes at the threshold, because Section 230 of the Communications Decency Act means a service is not treated as the publisher of what its users post. The rule is explained on what is Section 230 and applied to reviews on Section 230 and online reviews.

The consequence is simple to state and hard to accept: the claim runs against the person who typed it, not the site that displayed it. What the site will do is enforce its own policies, which is a separate and much faster process that has nothing to do with liability.

Review defamation and social media defamation

Reviews are the most common form. A review that evaluates a service is opinion and is protected. A review that states a false fact about what happened is not, and the difference is often a single sentence inside an otherwise lawful review.

Social media adds a republication question. At common law, each person who repeats a defamatory statement can be liable for it. Online, courts have applied Section 230 to individual users who forward or repost material written by someone else, though the boundaries are contested where the person adds their own commentary or alters what they are sharing. Someone who edits a quotation, or writes a false caption over a real screenshot, is the author of what they added.

Anonymous posters and the John Doe subpoena

Many damaging posts come from accounts with no real name attached. The route is to file a claim against a Doe defendant and seek a subpoena to the platform, and then often a second subpoena to an internet provider, to identify the account holder.

Courts do not grant this automatically. Most apply a threshold test first, requiring the claimant to show the claim has real substance before anonymity is stripped, precisely because anonymous speech is constitutionally protected. EFF on anonymous speech and unmasking subpoenas sets out the standards courts use and why they exist. In practice the process is slow, it is expensive, it becomes a public court record, and it sometimes ends with an account that cannot be traced to a person at all.

The clock, and the fee risk

Two features of the system catch people out.

Defamation limitation periods are short, commonly one to three years and in some states one, and the single publication rule means a post you found last week may have used up most of its clock years ago. Editing a page can sometimes restart it, but that is a state by state question.

The second is anti-SLAPP. Most states now have a statute allowing a defendant to seek early dismissal of a claim that targets protected speech, often with a fee award against the claimant. Filing a weak defamation claim over a genuine opinion is not a neutral act. It can end with you paying the other side.

What online defamation is not

• A truthful negative review, however much business it costs.
• An insult, a rant, or name calling with no checkable factual assertion.
• An accurate report of an arrest, a filing or a court record.
• A competitor's advertising claim, which is handled by advertising law rather than defamation.
• An autocomplete suggestion, in most cases. Google's own explanation of how Google Autocomplete predictions work describes them as generated from real searches rather than composed as statements, and claims aimed at the prediction itself have generally not succeeded in United States courts. Autocomplete is covered separately on Google autocomplete.

What to do first

Preserve the content before you do anything else: full page capture, visible URL, visible date. Then classify it, sentence by sentence, into factual assertions and opinion. Then check the platform's published policy, because a policy violation is removed free and quickly and does not require a legal theory. Only after those three steps is a lawyer's time well spent, and how to prove online defamation covers what they will ask you for.

Nothing on this page is legal advice, and the answer in a real case turns on the law of your state. The statutes, the case law and the platform guidance are linked so you can read the primary sources rather than trust a summary. Whether your problem is a defamation claim, a policy complaint, or a search result that simply needs to be outranked is the question worth settling first, and it is the first thing a reputation audit establishes.