What is “hate speech” and is it protected? A legal primer

What people mean by “hate speech” and why the term matters

In everyday use, “hate speech” describes words or messages that attack a person or group on the basis of race, religion, sex, sexual orientation, national origin, or other protected traits. This ordinary definition covers insults, slurs, slogans, and calls for exclusion or discrimination, but it is not a legal term of art in U.S. constitutional law.

Legally, the question is narrower: can the government punish that expression without violating the First Amendment. Private responses such as removal from a social platform or employer discipline are distinct from government suppression, and readers should separate social or civil consequences from constitutional punishment.

For legal standards, the controlling tests come from the Supreme Court and are discussed in the sections below, which summarize where the law draws lines around advocacy, threats, and harassment. For a general nonprofit explanation of how nonstate actors respond to hateful expression, see the ACLU explanation and resources ACLU explanation and resources. See the constitutional rights hub for related content on constitutional limits.

The core test in U.S. law: Brandenburg v. Ohio and imminent lawless action

The leading rule for when advocacy loses First Amendment protection asks two things: was the speech aimed at producing imminent lawless action, and was it likely to produce that action. That two-part standard comes from the Supreme Court decision in Brandenburg v. Ohio, which remains the primary limit on criminal punishment for advocacy of illegal acts Brandenburg v. Ohio (see the opinion on Justia Brandenburg v. Ohio opinion).

Put simply, advocacy of violence or illegality is not enough on its own to permit government punishment. The advocacy must be directed to producing immediate unlawful conduct, not to abstract arguments or remote encouragement, and the facts must show a real likelihood of imminent lawless action.

The Brandenburg standard explained, 2nd amendment bill of rights

Brandenburg sets an intent and imminence test. Courts look for words or conduct that clearly aim to trigger unlawful acts at once. A generalized call to violence at some unspecified future date typically remains protected, while a targeted, time and place specific call may fail constitutional scrutiny. See the LII Wex entry on the Brandenburg test Brandenburg test.

How courts assess direction and likelihood of imminent lawless action

Judges consider context, the speaker’s audience, and surrounding facts when deciding whether speech is both directed to and likely to produce imminent lawless action. Courts assess whether listeners were likely to act immediately and whether the speaker’s message was calibrated to produce that immediate reaction. A useful case summary appears on Oyez Brandenburg v. Ohio at Oyez.

Other established exceptions: fighting words, true threats, and intimidation

Fighting words and Chaplinsky

One limited category the Court has recognized is “fighting words,” described in Chaplinsky v. New Hampshire as words likely to provoke an immediate violent response from a hearer. The Court has since treated fighting words narrowly, so most offensive or hateful insults do not meet the high threshold for this exception Chaplinsky v. New Hampshire.

True threats and Elonis

Threats that place a person in fear for their safety can be unprotected, but modern doctrine requires attention to the speaker’s mental state. Elonis v. United States emphasized that courts must take account of intent or a culpable mental state when deciding criminal liability for threats, so context and state of mind are central to whether a statement counts as a punishable true threat Elonis v. United States.

Cross-context intimidation and Virginia v. Black

The Court has also recognized that conduct or speech used as intimidation in a particular context can be treated differently, especially when it targets groups with a history of racial intimidation. The decision in Virginia v. Black explores how cross-context factors and intent influence whether symbolic acts or words are punishable as intimidation or incitement Virginia v. Black.

When the government cannot punish but others still can: platforms, employers, and civil claims

Even when the First Amendment bars government punishment, private actors and civil law can produce consequences. Social platforms enforce terms of service and may remove or downrank hateful content under their rules, and employers may discipline employees under workplace policies for speech that affects operations or safety. Our freedom of expression and social media page discusses platform moderation issues.

Civil claims such as harassment suits, torts for intentional infliction of emotional distress, or other noncriminal remedies can also arise from targeted or threatening conduct, depending on state law and facts. For discussion of how civil and private responses differ from constitutional limits, the ACLU has a general overview of hate speech and First Amendment issues ACLU explanation and resources.

These nonstate mechanisms are active policy areas and can have significant practical effects even when government action would be unconstitutional.

How other countries treat hateful expression: a brief international comparison

U.S. constitutional law protects a broader range of hateful expression than many European frameworks, where criminal laws and international standards sometimes allow more extensive regulation of speech that incites hatred or discrimination. The gap means the same message may be punishable in one country but protected in U.S. courts.

For a practical overview of European approaches to regulating hateful expression, see the Council of Europe materials on combating hate speech Combating hate speech.

Jurisdiction matters for enforceability and choice of law, and cross-border online speech often raises complex questions about which rules apply and how they can be enforced in practice.

Applying classic tests to online speech and new technologies

Online publication, social sharing, and rapid viral spread complicate classic legal tests that were developed in a pre-internet era. The notion of imminence in Brandenburg is difficult to map when a message can be seen worldwide and amplified over hours or days rather than producing immediate local action.

AI-generated content and anonymity raise additional questions about attribution and intent. If an algorithm or a third party produces a threatening message, courts and platforms must sort out who was responsible and whether the mental state element required in cases like Elonis is present. For a general sense of how nonstate actors manage content, the ACLU discussion on platform moderation is useful ACLU explanation and resources.

Observers often conflate private moderation or employer discipline with government censorship. A social platform deleting a post is not the same as a criminal prosecution or a government order, and distinguishing the actor who acted is the first step in evaluating a claim.

Another frequent mistake is assuming offensive slogans or insults automatically cross into criminal liability. Under Brandenburg, Chaplinsky, and related decisions, courts require specific elements like imminence, targeted threats, or intent to intimidate before permitting criminal punishment; these elements keep many offensive expressions protected Brandenburg v. Ohio.

Use a short checklist when you evaluate a contested statement: who made it, to whom, what exactly was said, was there a time and place that made immediate action probable, was there apparent intent to intimidate or threaten, and which forum applied. Check primary cases and reputable explainers before concluding that speech was illegal.

In U.S. law, hateful ideas and offensive advocacy are generally protected, but narrow exceptions allow punishment for incitement to imminent lawless action, true threats, and a limited set of fighting words.

The Brandenburg test remains the principal limitation on government criminalization of advocacy Brandenburg v. Ohio.

Nonstate actors such as platforms and employers, and civil law claims, can impose consequences in ways the Constitution does not. For international contrasts, European bodies often permit broader regulation of hateful expression, so jurisdiction matters for punishability Combating hate speech.

Primary Supreme Court cases and reputable explainers are the best sources for firm answers on particular facts, and readers should consult those opinions when evaluating specific claims about protection and punishment. See our First Amendment explainer for background.