The goal is to provide clear, neutral guidance for readers who want to know when speech may lose constitutional protection and where to look for primary sources.
What counts as unprotected speech under the First Amendment
The Supreme Court has identified a short list of speech categories that receive no First Amendment protection. Those categories include incitement to imminent lawless action, true threats, obscenity, child sexual exploitation material, libel, and fighting words. Courts treat this list as narrow and tied to specific doctrinal tests rather than as a broad exception to free expression.
The identification of these categories stems from a series of Supreme Court opinions that set the governing rules for each area. For an entry point to the doctrine on incitement, see the Brandenburg opinion.
The legal categories named above come from separate lines of cases. Obscenity doctrine traces to a three-part test, libel law to a public-figure standard, fighting words to a breach-of-the-peace concept, threats to separate doctrines including cross-burning decisions, and child sexual exploitation material to criminal statutes the Court has upheld.
Because the Court limits unprotected speech to a few narrow categories, judges apply defined tests and fact-based inquiry before removing constitutional protection from particular expression. That approach leaves many offensive or provocative statements protected unless they meet those tests.
ToolType: | Purpose: | Fields: | Notes:
The governing legal tests the Court uses
Brandenburg test for incitement
The key test for incitement asks whether speech is directed to producing imminent lawless action and is likely to produce that action. The controlling case sets a high bar so that mere advocacy of ideas or even inflammatory political rhetoric is typically protected unless it aims at and is likely to cause immediate unlawful conduct.
Read the primary cases that define unprotected speech
The best source for working through these doctrines is the underlying Supreme Court opinions listed in this article. Reading the cases helps clarify how courts frame each test.
Miller test for obscenity
Obscenity is judged under a three-part test. Courts ask whether the work, taken as a whole, appeals to prurient interest under contemporary community standards, whether it depicts sexual conduct in a patently offensive way as defined by statute, and whether it lacks serious literary, artistic, political, or scientific value. That structure focuses on context and statutory definitions rather than broad moral disapproval.
Actual malice standard for libel
For libel involving public officials or public figures, the Supreme Court requires proof that a defamatory statement was made with actual malice, meaning the speaker knew the statement was false or acted with reckless disregard for the truth. This rule raises the evidentiary bar for successful defamation claims by public figures and shapes how courts balance reputation and free speech interests.
New York Times Co. v. Sullivan opinion
Chaplinsky on fighting words and Virginia v. Black on threats
Fighting words are a narrow category of face-to-face insults likely to provoke an immediate violent response; the concept is limited and rarely applied outside its specific factual setting. Separately, the Court has addressed threatening conduct and symbolic acts in decisions that clarify when intimidation or cross-burning is a true threat rather than protected expression.
Chaplinsky v. New Hampshire opinion
United States v. Williams on child sexual exploitation material
The Court has upheld criminal laws aimed at child sexual exploitation material, treating that category as unprotected because of the state’s interest in preventing the sexual abuse of minors and the distribution of materials depicting exploitation.
United States v. Williams opinion
How the incitement and true threat doctrines work in practice, including online speech
When courts analyze incitement claims they look for three elements: intent, imminence, and likelihood. Intent refers to whether the speaker meant to produce lawless action. Imminence asks whether the action urged is to occur right away. Likelihood examines whether the speech was likely to cause the action under the circumstances.
True threats are evaluated differently. Courts consider whether a reasonable listener would interpret the statement as a serious expression of intent to harm. The analysis distinguishes threats from political hyperbole by focusing on context, the speaker’s conduct, and whether the statement conveys a real intent to cause harm.
The Court treats a small set of categories as unprotected: incitement to imminent lawless action, true threats, obscenity, child sexual exploitation material, libel under strict standards when applicable, and fighting words, each governed by specific doctrinal tests.
Applying these standards to online speech raises practical challenges. Expressive acts shared on social media can reach wide audiences instantly, creating questions about who the intended audience is and whether a statement was likely to cause imminent lawless action in a specific place or time.
The digital context often requires courts to parse timing, platform mechanics, and the speaker’s online conduct. Lower courts continue to work through these factual puzzles, so many platform and algorithm issues remain contested in 2026.
Where ‘hate speech’ fits: protected expression versus narrow exceptions (hate speech first amendment)
The Supreme Court has not created a standalone exception labeled hate speech. Instead, hateful or offensive language is usually evaluated under existing categories like incitement, true threats, fighting words, or defamation to determine whether it loses First Amendment protection.
When hateful expression is targeted to produce imminent lawless action and meets the Brandenburg test, it loses protection for the same reasons any other incitement would. Conversely, harassment that amounts to a serious threat can be regulated as a true threat if it meets the legal criteria.
Hateful statements may also be treated as fighting words when they are face-to-face provocations likely to trigger immediate violence, though courts have applied that doctrine narrowly and with caution to avoid silencing political debate.
Chaplinsky v. New Hampshire opinion
For journalists and platform moderators the practical implication is that offensive content is not automatically outside First Amendment protection. Each claim requires fact-specific analysis to see whether the expression meets an existing exception under settled tests.
Readers should check primary sources or court rulings before concluding that speech is unprotected, because labeling content as hate speech does not by itself satisfy the tests courts use to remove constitutional protection.
Decision criteria for courts, platforms, and citizens when judging speech
Courts follow concrete decision factors for each category. For incitement they examine intent, imminence, and likelihood. For obscenity they evaluate community standards, statutory definitions, and whether the work lacks serious value. For libel they look for falsity and, in public-figure cases, actual malice. For fighting words they consider provocation and the risk of breach of the peace.
Private platforms operate under different constraints. A platform may remove or limit content under its own rules even when the speech would be legally protected. Platform enforcement is a private policy decision and does not equal a judicial finding that speech is unprotected under the First Amendment.
New York Times Co. v. Sullivan opinion
Citizens and journalists should treat inflammatory posts with caution. Before asserting that speech is unprotected, look for primary sources such as court opinions, statutes, or official findings that apply the legal tests to the specific facts at issue.
When in doubt, identify which legal test could apply and then seek the controlling case on that test to assess whether the threshold for unprotected status is met. For a site overview, see this discussion on unprotected speech.
Common mistakes and misconceptions about unprotected speech
A frequent error is assuming that offensive or hateful language is automatically unprotected. The law distinguishes offensive content from speech that meets the specific legal criteria for unprotected categories.
Another misconception is equating platform moderation with a legal ruling. Private removals reflect policy choices, not constitutional adjudication. Platforms can and do enforce standards that go beyond what the First Amendment requires.
New York Times Co. v. Sullivan opinion
People also misapply community standards or the actual malice rule, treating a single case as a broad template. Legal outcomes depend on detailed facts and the specific test that a court applies to those facts.
Practical examples and scenarios mapped to the legal tests
1) Online call to immediate violence: A post urges a crowd to storm a specific government building right now and supplies location details. Under the Brandenburg framework, courts would examine whether the speaker intended immediate lawless action and whether the post was likely to produce that action. If those elements are present a court could treat the post as unprotected incitement.
2) Obscene material shared on a public forum: An image gallery is evaluated under the Miller test. Courts would consider community standards where the material is posted, whether statutory definitions of sexual conduct are met, and whether the work lacks serious literary, artistic, political, or scientific value. The outcome can hinge on jurisdiction and context.
3) A threatening post aimed at a person: A social-media message telling a named individual that they will be physically harmed may be analyzed as a true threat. Courts ask whether a reasonable recipient would view the message as a serious expression of intent to harm, taking into account context and corroborating conduct.
4) Alleged libel about a public official: A widely circulated claim accusing a public official of criminal conduct will be judged under defamation law. If the official is a public figure, the plaintiff must prove falsity and actual malice, showing the speaker knew the statement was false or recklessly disregarded the truth.
New York Times Co. v. Sullivan opinion
Each hypothetical shows how context and factual detail determine whether similar words are protected or unprotected. Courts resolve real cases through fact finding, so simplified examples should be treated as illustrations, not legal rulings.
Conclusion: How to read claims that speech is not protected
The Supreme Court limits unprotected speech to a few narrow categories and ties each category to a specific legal test. That structure means context, intent, timing, and the presence of certain objective elements matter more than the offensiveness of the words alone.
Before concluding that speech is not protected, consult the cited cases and primary sources to see whether the facts align with the test that applies to the claimed category. Applying these doctrines to modern communications remains a live legal issue.
No. The Supreme Court has not created a standalone hate speech exception. Hateful language is often protected unless it meets a recognized unprotected category such as incitement, a true threat, fighting words, or defamation.
Online posts can be treated as incitement if they are intended to produce imminent lawless action and are likely to produce that action under the Brandenburg test, which requires intent, imminence, and likelihood.
No. Private platforms can remove content under their own rules even when the content remains legally protected; removal is a policy decision, not a judicial determination of unprotected speech.
Readers who want to explore further should read the controlling cases cited in this article to see how courts apply the tests to real facts.
References
- https://www.supremecourt.gov
- https://www.law.cornell.edu/supremecourt/text/395/444
- https://michaelcarbonara.com/issue/constitutional-rights/
- https://www.law.cornell.edu/supremecourt/text/413/15
- https://www.law.cornell.edu/supremecourt/text/376/254
- https://www.law.cornell.edu/supremecourt/text/315/568
- https://www.law.cornell.edu/supremecourt/text/553/285
- https://www.law.cornell.edu/supremecourt/text/538/343
- https://docs.fcc.gov/public/attachments/DOC-310963A1.pdf
- https://michaelcarbonara.com/freedom-of-expression-and-social-media/
- https://www.congress.gov/crs-product/IF11072
- https://michaelcarbonara.com/contact/
- https://michaelcarbonara.com/bill-of-rights-first-amendment-unprotected-speech/
