Freedom of speech in the United States has real limits, but they are narrower and more technical than most people assume. The First Amendment stops the government from punishing the vast majority of what you say, write, or express, and it protects offensive, unpopular, and even hateful ideas. What it does not shield is a short list of speech categories the Supreme Court has carved out over the past century, along with reasonable rules about when and where speech happens. Understanding the actual limits on freedom of speech means knowing that short list, knowing who the First Amendment binds, and knowing what does not count as a limit at all.
Who the First Amendment Actually Binds
The First Amendment restricts the government. Federal, state, and local officials, public agencies, and publicly funded institutions cannot punish you for protected expression or write laws that silence viewpoints they dislike.1LII / Legal Information Institute. State Action Doctrine and Free Speech
Private companies are not bound by it. A social media platform can remove your posts. A private employer can fire you over something you said at work. A shopping mall can eject you for handing out flyers. None of that raises a First Amendment issue, because no government actor is involved.
There is one narrow exception. When a private entity performs a function that has traditionally and exclusively belonged to the government, courts can treat it as a state actor. The Supreme Court applied this reasoning to a company-owned town that functioned like a municipality. The exception is very hard to trigger, and running a business open to the public is not enough.2LII / Legal Information Institute. State Action Doctrine
Categories of Speech the Government Can Punish
The Supreme Court has identified a small set of speech categories that receive no First Amendment protection. Each reflects a judgment that the speech causes concrete harm and adds little to public debate. Outside these categories, government efforts to punish speech based on its content almost always fail.
Incitement to Imminent Lawless Action
The government can punish speech aimed at sparking immediate illegal conduct, but only under a demanding two-part test. The speaker must be intentionally pushing the audience toward imminent lawless action, and the speech must be genuinely likely to produce that result.3Justia U.S. Supreme Court Center. Brandenburg v. Ohio, 395 U.S. 444 (1969)
Both prongs matter. Urging an angry crowd to storm a building right now can meet the test. A fiery speech abstractly endorsing revolution someday does not. Vague calls for future lawbreaking, however disturbing, stay protected.
True Threats
A true threat is a statement communicating a serious intent to commit violence against a specific person or group. In 2023, the Supreme Court held that prosecutors must prove the speaker acted with at least recklessness, meaning the speaker consciously disregarded a substantial risk that the words would be understood as a real threat of violence.4Supreme Court of the United States. Counterman v. Colorado
Context does most of the work. Political hyperbole, dark jokes, and heated rhetoric can all sound threatening on the page. Courts examine the full circumstances to sort a genuine threat from language a reasonable listener would not take literally.
Fighting Words
Fighting words are face-to-face insults so provocative that they are likely to trigger an immediate physical fight. The Supreme Court first defined them in 1942 as words that “by their very utterance inflict injury or tend to incite an immediate breach of the peace.”5Justia U.S. Supreme Court Center. Chaplinsky v. New Hampshire, 315 U.S. 568 (1942)
In practice, the Court has since narrowed the category to a direct personal insult or invitation to fight aimed at a specific individual.6Justia U.S. Supreme Court Center. Texas v. Johnson, 491 U.S. 397 (1989) Offensive signs, vulgar language in public, and insults aimed at a crowd have all been found to fall outside it. The category exists on paper but rarely succeeds in court.
Defamation
Defamation covers false statements of fact that damage someone’s reputation. Libel is written; slander is spoken. A plaintiff generally has to show a false factual statement, communication to someone other than the subject, some level of fault, and real reputational harm.
When the target is a public official or public figure, the bar is much higher. Under New York Times Co. v. Sullivan, a public figure can only win by proving the speaker either knew the statement was false or acted with reckless disregard for the truth.7Justia U.S. Supreme Court Center. New York Times Co. v. Sullivan, 376 U.S. 254 (1964) The Court set this “actual malice” standard to keep critics of government from self-censoring out of fear of unlimited damages.
Opinions, no matter how harsh, are not defamation. Calling a restaurant’s food terrible is a subjective judgment. Falsely claiming the restaurant serves expired ingredients, and doing measurable damage to its business, could be actionable. The line runs between provably false assertions of fact and everything else.
Obscenity
Obscenity is unprotected, but the legal definition is narrower than what most people would call obscene. The three-part Miller test asks whether a work, taken as a whole, appeals to a sexual interest under contemporary community standards; depicts sexual conduct in a way that is patently offensive under applicable state law; and lacks serious literary, artistic, political, or scientific value.8Justia U.S. Supreme Court Center. Miller v. California, 413 U.S. 15 (1973)
All three prongs must be met. Crude, vulgar, or sexually explicit material with genuine artistic or political value stays protected. The “serious value” prong is judged by a national standard, which is one reason successful obscenity prosecutions are rare.
Child Sexual Abuse Material
Material depicting the sexual exploitation of real children is unprotected as its own category, separate from general obscenity. In New York v. Ferber, the Supreme Court held that the government’s interest in protecting children is so compelling that this material can be banned even when it would not meet the Miller test.9Justia U.S. Supreme Court Center. New York v. Ferber, 458 U.S. 747 (1982) The line is drawn at the involvement of actual minors in the production of the material.10Congress.gov. Social Media – Regulatory, Legal, and Policy Considerations for the 119th Congress
Hate Speech Is Not on the List
There is no hate speech exception to the First Amendment. The phrase has no formal meaning in U.S. constitutional law, and the Supreme Court has repeatedly refused to create one. Speech that demeans people based on race, religion, gender, sexual orientation, or similar characteristics is protected unless it independently qualifies as a true threat, incitement, fighting words, or another unprotected category.
The Court has been direct on this. In Matal v. Tam, the justices wrote that speech cannot be banned simply because society finds the ideas offensive or disagreeable, calling the protection of hateful expression “the proudest boast of our free speech jurisprudence.”11LII / Legal Information Institute. Matal v. Tam In R.A.V. v. City of St. Paul, the Court struck down a local ordinance that singled out bias-motivated speech for punishment, holding that the government cannot pick and choose based on the speaker’s viewpoint even within otherwise regulable categories.12LII / Legal Information Institute. R.A.V. v. City of St. Paul
The rule holds even when the speech is deeply painful in context. In Snyder v. Phelps, the Court protected the Westboro Baptist Church’s protests near military funerals because they addressed matters of public concern in a public place without disruption. The emotional impact on the grieving family did not strip the speech of protection.13Justia U.S. Supreme Court Center. Snyder v. Phelps, 562 U.S. 443 (2011)
Prior Restraint: The Line Against Censorship Before It Happens
Even where the government has some ground to punish speech after the fact, blocking it in advance is a different matter. Prior restraint means the government tries to stop speech before it happens through licensing, an injunction, or outright censorship. Courts treat these attempts as presumptively unconstitutional and almost always strike them down.14Justia U.S. Supreme Court Center. Near v. Minnesota, 283 U.S. 697 (1931)
The Supreme Court acknowledged in Near v. Minnesota that the ban is not absolute. Publishing troop movements during wartime or distributing obscene material might justify stopping speech in advance. Outside those extreme cases, the remedy for harmful speech is punishment after publication, not censorship before it. This is why courts are so skeptical of gag orders, pre-publication injunctions, and permit schemes that hand officials discretion to say no.
Rules About When, Where, and How
Protected speech can still be regulated in terms of when, where, and how it happens. A city can require permits for large marches, enforce nighttime noise limits, and cap the size of signs on public property. Under Ward v. Rock Against Racism, such rules are constitutional if they meet three requirements:
- They apply to all speech regardless of its message.
- They are narrowly tailored to serve a significant government interest such as public safety or traffic flow.
- They leave open ample alternative channels for the message.
A ban on all amplified sound in a park after 10 p.m. passes easily. A ban on amplified sound only when the message criticizes the city does not, because it targets content.15Justia U.S. Supreme Court Center. Ward v. Rock Against Racism, 491 U.S. 781 (1989)
Where the speech occurs matters too. Sidewalks and parks are traditional public forums with the strongest protections; any content-based restriction there faces the most demanding scrutiny. On government property that has never been opened for public speech, such as a military base or a jail, officials have far more room, so long as the rules are reasonable and viewpoint-neutral.
Commercial Speech Gets Less Protection
Advertising and other speech that proposes a commercial transaction is protected, but less than political or personal expression. Consumers have a right to receive truthful information about legal products and services.16Justia U.S. Supreme Court Center. Virginia State Board of Pharmacy v. Virginia Citizens Consumer Council, 425 U.S. 748 (1976)
Commercial speech that is false, misleading, or promotes illegal activity can be banned outright. For truthful advertising about lawful products, the government must satisfy the four-part Central Hudson test: the speech must concern lawful activity and not be misleading; the government must identify a substantial interest; the regulation must directly advance that interest; and it must not be more extensive than necessary.17Justia U.S. Supreme Court Center. Central Hudson Gas and Electric Corp. v. Public Service Commission, 447 U.S. 557 (1980) This framework lets governments regulate tobacco advertising or misleading health claims while blocking rules aimed at commercial messages officials simply dislike.
Speech in Schools and Government Jobs
Two settings have their own rules that trip people up.
Students keep First Amendment rights inside public schools. The Court held in Tinker v. Des Moines that students do not “shed their constitutional rights to freedom of speech or expression at the schoolhouse gate.”18Justia U.S. Supreme Court Center. Tinker v. Des Moines Independent Community School District, 393 U.S. 503 (1969) Schools can restrict student speech that materially and substantially disrupts the educational environment or invades the rights of other students, but the disruption has to be concrete, not just administrator discomfort. For off-campus speech, including on social media, the Court has said schools must tread carefully, because that speech typically falls under parental supervision and regulating it both on and off campus could silence students entirely.19Justia U.S. Supreme Court Center. Mahanoy Area School District v. B.L., 594 U.S. ___ (2021)
Public employees are treated differently when they speak as part of their job. Under Garcetti v. Ceballos, the First Amendment “does not insulate their communications from employer discipline” when they are speaking in their professional capacity.20Justia U.S. Supreme Court Center. Garcetti v. Ceballos, 547 U.S. 410 (2006) When a government employee speaks as a private citizen on a matter of public concern, such as a teacher writing a letter to the newspaper about the school board’s budget, some protection applies, and courts balance the employee’s interest in speaking against the government’s interest in an efficient workplace.