Free Speech and Its Limits: What the First Amendment Actually Protects
Few constitutional provisions generate more confusion than the First Amendment’s free speech guarantee. It is genuinely broad — broader than in most other democracies — but it has defined limits, and knowing those limits is the starting point for any serious discussion of speech, censorship, or platform moderation.
Published June 28, 2026The First Amendment states that Congress shall make no law “abridging the freedom of speech, or of the press.” Through the Fourteenth Amendment, this restriction applies to state and local governments as well. The protection is one of the most expansive in the world, covering political speech, artistic expression, commercial advertising, and much else — including a great deal of speech that many people find offensive or harmful.
But the amendment begins with “Congress shall make no law.” This matters. The First Amendment restricts government action, not private action. A private employer, social media platform, or university can restrict speech on their own premises or services without implicating the First Amendment, because they are not government actors. Confusion between what the government cannot do and what private institutions choose to do is one of the most common misconceptions in everyday speech-related debates.
Categories of unprotected speech
The Supreme Court has identified a limited set of speech categories that fall outside First Amendment protection. These are not vague exceptions that courts apply broadly; they are defined categories with specific legal standards.
Incitement to imminent lawless action. The governing standard comes from Brandenburg v. Ohio (1969), which held that government can only punish speech that is directed to inciting or producing imminent lawless action and is likely to produce such action. Abstract advocacy of illegal activity or violence at some unspecified future time is protected; specific calls for immediate criminal action against identifiable targets are not. This is a narrow standard that protects a great deal of inflammatory speech.
True threats. Statements that communicate a genuine, serious intention to commit violence against a specific person or group are not protected. Courts look at whether a reasonable person would interpret the communication as a threat, not just whether the speaker claims they did not mean it literally.
Defamation. False statements of fact that damage someone’s reputation can be actionable. Public figures face a higher standard: they must show that the speaker acted with “actual malice” — knowing the statement was false or acting with reckless disregard for its truth or falsity. Private individuals face a lower bar but still must show some level of fault.
Obscenity. Sexually explicit material that meets the legal definition of obscenity under the three-part Miller test can be regulated. The test considers whether the material appeals to prurient interest, whether it depicts sexual conduct in a patently offensive way, and whether it lacks serious literary, artistic, political, or scientific value. Material that does not satisfy all three parts of the test is not legally obscene, even if many people find it objectionable. Child sexual abuse material is categorically unprotected by a separate and more absolute standard.
Fraud and perjury. Deliberately false statements made in contexts like contracts, sworn testimony, or certain commercial transactions can be regulated without implicating free speech protections in the ordinary sense.
What is clearly protected
Outside these narrow categories, the government’s ability to restrict speech based on its content is extremely limited. Political speech receives the highest level of protection, even when it is offensive, inflammatory, or deeply objectionable to large segments of the public. The government cannot ban speech simply because it disagrees with the viewpoint expressed or because the speech makes people uncomfortable.
Hate speech — speech that expresses hostility toward people based on characteristics like race, religion, or sexual orientation — is broadly protected in the United States. This is a significant difference from many other democracies, which have enacted hate speech laws. The American approach reflects a judgment that the risks of giving government authority to define and punish offensive ideas outweigh the harms of the speech itself, and that countering bad speech with more speech is preferable to legal suppression. This remains a genuinely contested policy question; the constitutional answer and the policy question are separate issues.
Content-neutral regulations
The government has more room to regulate the time, place, and manner of speech than to regulate its content. A city can require a permit for a large demonstration in a park, impose reasonable noise limits, or designate specific areas for protests near government buildings — as long as these rules apply equally to all speakers regardless of their message and leave open alternative channels for communication. What the government cannot do is apply those rules selectively to suppress disfavored viewpoints.
Why this matters
Speech debates in contemporary public life frequently involve private platforms, universities, and employers rather than government. The legal First Amendment framework does not resolve those disputes directly. But the values underlying the First Amendment — the importance of open discourse, the danger of letting authorities decide which ideas are acceptable, the role of speech in democratic self-governance — inform those debates even when they are not legally controlling. Understanding what the law actually says is the necessary starting point for thinking clearly about what policies, public or private, are wise.