The First Amendment protects most speech, but not all of it. Courts have identified a short, tightly defined list of First Amendment exceptions: incitement to imminent lawless action, true threats, fighting words, obscenity, child sexual abuse material, defamation, and speech that is itself a crime such as fraud, perjury, or bribery. Commercial speech gets less protection than political or artistic expression but is not unprotected. And before any of that matters, one threshold rule applies: the First Amendment restricts the government, not private employers, platforms, or schools.
The First Amendment Only Limits Government Action
The text says “Congress shall make no law…abridging the freedom of speech,”1National Archives. The Bill of Rights: A Transcription and through the Fourteenth Amendment that prohibition reaches state and local governments too. The operative word is government. A private company firing an employee over a social media post, a platform removing content that violates its terms of service, or a private university disciplining a student is not committing a First Amendment violation.
The Supreme Court reinforced this in Manhattan Community Access Corp. v. Halleck, holding that a private entity designated by a city to run public access television channels was not a government actor bound by the First Amendment.2Justia. Manhattan Community Access Corp. v. Halleck, 587 U.S. ___ (2019) Merely providing a forum for speech does not turn a private entity into the government. So when someone says a social media ban violated their First Amendment rights, they are almost always wrong as a matter of law.
Incitement to Imminent Lawless Action
Speech that pushes people toward immediate illegal activity can be punished, but only under a demanding standard. In Brandenburg v. Ohio (1969), a case involving a Ku Klux Klan leader, the Supreme Court held that speech loses protection only when it is both directed at producing imminent lawless action and likely to succeed in doing so.3Cornell Law Institute. Brandenburg Test Both elements have to be present. Urging a crowd already gathered outside a building to break down the doors right now crosses the line. Posting an abstract manifesto calling for revolution “someday” does not.
Inflammatory rhetoric, calls for general resistance, and even disturbing advocacy of illegal ideas remain protected as long as they lack that direct, immediate link to likely criminal action. The bar is high on purpose. A lower one would let the government punish dissent whenever it found the message dangerous.
True Threats and Fighting Words
A true threat is a serious expression of intent to commit violence against a specific person or group. The Supreme Court defined the category in Virginia v. Black, explaining that true threats lose protection because of the fear they inflict on their targets, regardless of whether the speaker actually plans to follow through.4Cornell Law Institute. Virginia v. Black That is different from political hyperbole. In Watts v. United States, the Court treated a protester’s crude statement about the president as overheated political rhetoric and therefore protected speech.5Cornell Law Institute. Watts v. United States, 394 U.S. 705 Context is what tells the two apart.
In 2023, the Court added a mental-state requirement. Counterman v. Colorado held that prosecutors must prove the speaker had at least a reckless awareness that their statements would be perceived as threatening, meaning the person consciously disregarded a substantial risk that their words would cause fear.6Supreme Court of the United States. Counterman v. Colorado (2023) Before Counterman, courts were split on whether the speaker’s state of mind mattered at all. The recklessness standard protects people who genuinely do not realize their words sound threatening while still allowing prosecution of those who know exactly what they are doing.
Fighting words cover related but narrower ground. In Chaplinsky v. New Hampshire (1942), the Supreme Court described them as face-to-face insults so personally abusive that they would provoke a reasonable person to an immediate violent reaction.7Library of Congress. U.S. Reports: Chaplinsky v. New Hampshire, 315 U.S. 568 (1942) Courts have narrowed the category significantly since then. Speech that is merely offensive, vulgar, or enraging to a general audience does not qualify. The confrontation has to be direct, personal, and likely to produce an immediate physical altercation.
Obscenity and Child Sexual Abuse Material
Obscenity has been unprotected since the Court’s earliest free-speech cases. The modern test comes from Miller v. California (1973). Material is legally obscene only if it meets all three of these conditions: the average person, applying local community standards, would find that the work appeals to a prurient interest in sex; the work depicts sexual conduct in a way that is patently offensive under the law; and the work, taken as a whole, lacks serious literary, artistic, political, or scientific value.8Cornell Law School. Overview of Obscene Speech That third prong is what keeps obscenity law from sliding into censorship, protecting work with genuine creative or intellectual merit even when some people find it offensive.
Child sexual abuse material is a separate category with a separate rule. In New York v. Ferber (1982), the Court held that CSAM does not need to meet the obscenity test at all. Because real children are harmed in its production, the government can ban its creation, distribution, and possession outright, regardless of any claim to artistic value.9Justia. New York v. Ferber, 458 U.S. 747 (1982) Federal penalties are severe. A first conviction for distributing CSAM carries a mandatory minimum of five years and a maximum of twenty years in federal prison. A repeat offender faces a mandatory minimum of fifteen years and up to forty.10Office of the Law Revision Counsel. 18 USC 2252 – Certain Activities Relating to Material Involving the Sexual Exploitation of Minors
Defamation
A false statement of fact that damages someone’s reputation is not protected. Defamation divides into libel (written) and slander (spoken) and is primarily a state-law claim pursued in civil court. A plaintiff generally has to show that the defendant communicated a false factual statement to someone other than the plaintiff and that the statement caused real harm.
Who was defamed changes the standard. Public officials and public figures must meet the “actual malice” standard from New York Times Co. v. Sullivan (1964), proving the speaker either knew the statement was false or acted with reckless disregard for its truth.11Library of Congress. New York Times Co. v. Sullivan, 376 U.S. 254 (1964) That is an intentionally tough bar, and it exists to protect robust public debate even when journalists or critics get facts wrong, so long as they were not deliberately lying or willfully ignoring the truth. Private individuals face a lower threshold, typically negligence.
Fraud and Speech That Carries Out a Crime
When words are the mechanism for committing a crime rather than the expression of an idea, the First Amendment offers no shelter. Fraud, perjury, bribery, and extortion are the clearest examples. In each, the speech is the crime, not commentary about the crime.
Federal penalties reflect that. Perjury, lying under oath in a legal proceeding, carries up to five years in federal prison.12Office of the Law Revision Counsel. 18 U.S. Code 1621 – Perjury Generally Bribery of a public official can result in up to fifteen years.13Office of the Law Revision Counsel. 18 U.S. Code 201 – Bribery of Public Officials and Witnesses Fraudulent commercial speech, meaning a business making false claims to deceive consumers, triggers enforcement by the Federal Trade Commission, which can seek court orders freezing assets, stopping the fraud, and securing compensation for victims.
The unifying principle is criminal intent. The government is not punishing the speaker’s viewpoint. It is punishing the use of words as an instrument of theft, corruption, or deception.
Commercial Speech Gets Less Protection
Advertising and other commercial speech sit in a middle tier. They are not stripped of protection like obscenity or true threats, but they receive less than political or artistic expression. The framework comes from Central Hudson Gas & Electric Corp. v. Public Service Commission (1980). The speech must concern lawful activity and not be misleading. If it clears that step, any government regulation must serve a substantial interest, directly advance that interest, and be no broader than necessary.14Cornell Law Institute. Commercial Speech
The government can ban outright false advertising but cannot prohibit truthful ads simply because it dislikes the product. A state might restrict how tobacco companies market to children. It could not ban all alcohol advertising just because some legislators think drinking is harmful.
There Is No Hate Speech Exception
Hate speech, on its own, is not one of the categories. This surprises many people, but the Supreme Court has been direct about it. In Matal v. Tam (2017), the justices struck down a federal law banning registration of “disparaging” trademarks, saying “speech may not be banned on the ground that it expresses ideas that offend.”15Justia. Matal v. Tam, 582 U.S. ___ (2017) In Snyder v. Phelps (2011), the Court held that Westboro Baptist Church’s deeply offensive picketing at military funerals was protected because it addressed matters of public concern.16Justia. Snyder v. Phelps, 562 U.S. 443 (2011)
Hateful speech can still carry legal consequences when it also qualifies as a true threat, incitement, or targeted harassment under existing law. What the government cannot do is punish speech simply because the ideas expressed are bigoted. Many other democracies take the opposite approach; U.S. law does not.
Time, Place, and Manner Rules Are Not Exceptions
Even fully protected speech can be subject to rules about when, where, and how it happens. A city can require a permit for a march through downtown, cap the decibel level of amplified speech in a residential neighborhood, or designate specific areas for demonstrations near a courthouse. These are not exceptions to the First Amendment. They are limits on logistics, not content.
To survive a legal challenge, such restrictions must be content-neutral, narrowly tailored to serve a significant government interest, and leave open alternative ways to communicate.17Constitution Annotated. Overview of Content-Based and Content-Neutral Regulation of Speech A noise ordinance that applies equally to protest rallies and block parties is fine. A noise ordinance that applies only to political demonstrations is not. The setting also matters. In traditional public forums like parks, sidewalks, and public squares, the government faces strict scrutiny for any content-based restriction. In nonpublic forums, officials have more leeway as long as the rules are reasonable and viewpoint-neutral.18Cornell Law Institute. Forums Courts watch closely for restrictions that look neutral on paper but single out specific speakers in practice.