What the First Amendment Actually Protects

Last updated 14 hours ago. Our resources are updated regularly but please keep in mind that links, programs, policies, and contact information do change.

Fire off one angry tweet and you may learn the hard way that the First Amendment will not save your job. Ask ten people what it protects and nine will say freedom of speech.

Here is the thing that trips people up: the First Amendment is not a general fairness code for arguments. It does not guarantee that your boss cannot fire you for a tweet, that Facebook has to keep your post up, or that a private university has to let you speak. It is a limit on government, and only government, except in a handful of narrow situations.

So if you want to know whether something you plan to say, post, or do is protected, the honest answer is that it depends on three questions. Who is trying to control your speech? What kind of speech is it? And how does the restriction work?

Answer those, line them up with a few key rules the Supreme Court has built over the last fifty years, and you can usually predict how a court would treat your situation before you ever call a lawyer.

Even people who know this get it wrong under pressure. Athar Haseebullah, a civil-rights lawyer interviewed by Reno’s News 4, put it plainly in a segment on speech at work and on private platforms. The First Amendment shields you from government punishment, not from the rules set by a private employer or company.

The Question Behind Every Free Speech Fight

The text itself gives you the first clue. It begins “Congress shall make no law,” and courts have long stretched that to cover every branch and every level of government, from Congress down to a city council. What it has never covered is a purely private decision about speech.

That distinction is the whole ballgame for everyday disputes. When the American Civil Liberties Union explains protest rights, it frames them specifically as protections against “police and other government officials.” Not against a shopping mall. Not against your employer.

Public understanding lags well behind that reality. A 2025 survey by the Annenberg Public Policy Center found that while 79 percent of respondents could name “freedom of speech” as a First Amendment right, fewer than half of people could name any of the others.

The second boundary is that even government cannot be stopped from punishing certain speech, because some speech does direct harm. The Supreme Court recognizes categories that get little or no protection: obscenity, true threats, and, with a lighter touch, commercial advertising.

Obscenity runs through the three-part test from Miller v. California, which defines obscenity as material whose predominant appeal, to the average person applying contemporary standards, is to prurient interest — long defined in the Court’s obscenity cases as “a shameful or morbid interest in nudity, sex, or excretion” — depicts sexual conduct in a patently offensive way, and lacks serious literary, artistic, political, or scientific value. Hate speech, for what it is worth, is not on that list. As we cover in our look at 5 First Amendment Myths That Could Get You in Trouble, offensive speech is generally protected; it is the narrow harm categories that are not.

Justice Oliver Wendell Holmes Jr. drew one of those lines a century ago in Schenck v. United States, using the image of falsely shouting fire in a theater to argue that protection does not reach speech creating a clear and present danger. The categories have shifted since, but the principle held: not every restriction is automatically off-limits.

With that frame set, walk through the situations people actually hit.

When the Government Leans on Platforms

Start with the most modern version of the problem. Your post gets removed. An official somewhere had complained about that kind of content. Is that a First Amendment violation?

Usually not, and here is why. Under the state action doctrine, the First Amendment restrains only government. So Facebook, X, and YouTube, as private companies, are free to set their own moderation rules without following constitutional speech rules, any more than a newspaper editor has to print your letter. Our earlier analysis of the amendment’s limits walks through it in more depth.

But there is a wrinkle. When government officials apply real pressure behind the scenes, or “significantly encourage” a private company, pushing it hard enough to make it censor, courts can sometimes treat the private decision as the government’s own. Scholars call this dynamic jawboning, and a growing body of legal scholarship examines exactly where informal government pressure crosses the line into unconstitutional censorship.

The line is between persuasion, which is fine, and coercion, which is not. That is where Murthy v. Missouri comes in.

Decided June 26, 2024, the case grew out of a sweeping lawsuit by Missouri, Louisiana, and several individuals against federal agencies including the Surgeon General, the CDC, and the FBI. The plaintiffs alleged, in the Court’s own summary, that Executive Branch officials and agencies… pressured the platforms to suppress their speech on COVID-19 and election integrity.

The Fifth Circuit had narrowed a broad district-court injunction but still found certain officials likely crossed into coercion. The Supreme Court did not reach that question at all. Writing for the majority, Justice Amy Coney Barrett held that “neither the individual nor the state plaintiffs have established Article III standing to seek an injunction against any defendant.” Standing is the legal right to bring the case at all.

So the Court threw out the lower ruling on a technical hurdle, without deciding whether the government actually broke the law. The plaintiffs claimed “direct censorship” of their speech, but the Court found the platforms had their own moderation policies and often removed content with no government input. The injuries were not clearly “traceable to the Government defendants.” In plain terms, it wasn’t clear the government had actually caused their harm.

The practical takeaway for you is blunt. If your account gets suspended, you generally cannot win a First Amendment suit unless you can show a government actor did not just flag your content but effectively dictated the result. You must also show the harm is ongoing and that an injunction would stop it. Without that record, your dispute is a private one, not a constitutional one.

Threats, Stalking, and the Recklessness Line

Aggressive speech is a different problem, because true threats fall outside protection entirely. The trick is separating a genuine threat from dark humor or a heated political slogan.

For years, courts split on what the speaker had to be thinking. Some asked only whether a reasonable person would feel threatened. Others demanded proof the speaker actually knew.

The Supreme Court settled it in Counterman v. Colorado, decided June 27, 2023.

Billy Raymond Counterman sent hundreds of unsolicited Facebook messages over several years to a local singer identified in the opinion as C.W. The messages implied he was watching her and kept coming after she blocked him. Colorado convicted him of stalking under an objective standard, asking only whether a reasonable person would suffer serious emotional distress, without any finding about his own state of mind.

Justice Elena Kagan, writing for the Court, reversed. The State, she held, must prove the defendant himself grasped that his words were threatening. But the Constitution requires no more demanding a showing than recklessness: a conscious disregard of a substantial risk that the words would be seen as a threat. The Court drew on Virginia v. Black, which had struck down a cross-burning law for lacking that subjective element.

Justice Sonia Sotomayor concurred in part, calling recklessness enough for this case while worrying it might not give threats enough “breathing space” in other contexts.

What this means for you: prosecutors cannot convict on threatening words alone. They have to show you consciously ignored the risk your message would land as a threat. Repeated contact after a warning, references to someone’s movements, language of surveillance, all of that is evidence of recklessness. An offhand violent metaphor at a rally is not.

Victim advocates warn the new floor makes protection harder to win, with one advocacy group publicly stating its disappointment in the decision. Mary Anne Franks, a professor of law at George Washington University and president of the Cyber Civil Rights Initiative, argues in the George Washington Law Review that objective, victim-centered standards better track real danger, since a stalker’s inner state is easy to deny and hard to prove.

The counterargument, offered in Justice Barrett’s dissent, is that most unprotected categories, from obscenity to fighting words, use objective tests without demanding proof of recklessness, so true threats should not get special constitutional treatment. A 2026 California Law Review analysis documents lower courts already reading Counterman to require proof of the speaker’s guilty state of mind in stalking-by-speech cases, with at least one court limiting it to prosecutions based on message content.

When a Business Can Refuse to Speak

The amendment protects your right not to speak, too. Government cannot force you to mouth a message you reject.

It struck down mandatory flag salutes in West Virginia State Board of Education v. Barnette and refused to make private groups carry messages they disowned in Hurley and Boy Scouts of America v. Dale.

303 Creative LLC v. Elenis, decided June 30, 2023, brought that line into the age of custom web design. Lorie Smith, who owns a Colorado design business, wanted to build wedding websites but not for same-sex weddings, citing her religious beliefs.

She feared Colorado’s Anti-Discrimination Act, which bars public-facing businesses from denying service based on sexual orientation, would force her hand. So she sued before Colorado had taken any action against her, even before any customer complaint, according to Oyez.

Justice Neil Gorsuch, writing for a six-justice majority, framed the question as whether Colorado could “force an individual to speak in ways that align with its views but defy her conscience about a matter of major significance.” Both sides had agreed that Smith’s sites would be expressive, pure speech, so the Court held the state could not compel her to create them.

The ACLU, in an amicus brief, argued the opposite: that applying the statute “does not violate the First Amendment” because it required equal treatment, not endorsement of any message.

The two lessons for you cut against each other. If your work is genuinely expressive, some anti-discrimination applications may count as compelled speech. But the ruling does not authorize blanket discrimination, and it turns entirely on how a court characterizes the service. Off-the-shelf goods are not custom expression.

How Advertising Rules Get Judged

Commercial speech, meaning speech that does little more than propose a transaction, is protected but held to a looser standard. The framework comes from Central Hudson Gas & Electric Corp. v. Public Service Commission, decided June 20, 1980.

New York had banned promotional advertising by electric utilities to save energy after the 1970s crisis. The Court struck it down. As the opinion put it, “the First Amendment protects commercial speech from unwarranted governmental regulation,” and a total ban was more extensive than the conservation goal required.

Here is what each part means in practice.

Source: a hosted copy of Central Hudson v. Public Service Commission (1980).

The test explains a lot of everyday law. Linmark Associates v. Township of Willingboro used it to strike down a ban on “For Sale” signs meant to slow white flight.

Zauderer v. Office of Disciplinary Counsel upheld a rule making lawyers disclose that clients might owe costs, because factual disclosures aimed at preventing deception are far less constitutionally troublesome than restrictions on political speech.

That intermediate treatment is contested. A listener-focused view, in a law review article titled “Commercial Speech Protection as Consumer Protection,” holds that protection is “justified principally by the value to consumers of the information such speech provides.” On that reasoning, rules that curb manipulation without cutting real information fit the First Amendment rather than fighting it.

Public-health regulators press the point with data. A meta-analysis in Tobacco Control found advertising bans linked to roughly 20 percent lower odds of current smoking and a 37 percent reduced risk of initiation, evidence public-health regulators say satisfies the Central Hudson test’s demand that a rule “directly advance” a real interest.

Why Campaign Spending Counts as Speech

Few areas confuse people more than campaign finance, because the law talks about money using the language of speech.

It starts with Buckley v. Valeo, decided January 30, 1976. The Court’s unsigned opinion drew the line that still structures everything: “The Act’s contribution provisions are constitutional, but the expenditure provisions violate the First Amendment.” Limits on what you give directly to a candidate can stand, because they fight corruption. Limits on what you spend to promote your own views cannot.

The justices were direct about why: “The First Amendment denies government the power to determine that spending to promote one’s political views is wasteful, excessive, or unwise.” It also upheld disclosure requirements and the public financing of presidential campaigns as constitutional.

Citizens United v. FEC, decided in 2010, extended that logic. The case began when a conservative nonprofit challenged FEC rules blocking it from broadcasting a film critical of then-candidate Hillary Clinton near a primary.

Justice Anthony Kennedy, writing for the majority, held that limits on independent spending by corporations and unions are limits on speech. Combined with the lower-court decision in Speechnow.org v. FEC, it gave rise to super PACs, which the FEC describes as committees that “may accept unlimited contributions, including from corporations and labor organizations.”

Whether that distinction holds up in the real world is the heart of the reform critique. Ciara Torres-Spelliscy, a professor of law at Stetson University and former Brennan Center fellow, laid out the numbers in congressional testimony titled “Taking Stock of Citizens United v. FEC”. Outside groups spent nearly $4.5 billion in the decade after the ruling, against about $750 million over the prior two decades.

Her testimony traces the concentration further. In 2016, more than 2,300 super PACs spent about $1.1 billion, most of it from just 100 donors. State systems followed: Colorado’s independent spending rose from under $400,000 in 2006 to about $136.9 million by 2018. Torres-Spelliscy argues the corporate money behind political ads is often hard to trace at all.

The doctrine is settled; the argument over it is not.

What Government Workers Can and Cannot Say

Work for the government and you get some First Amendment protection your private-sector neighbor does not. It is qualified, and it turns on whether you were speaking as a citizen or doing your job.

The foundation is Pickering v. Board of Education, decided in 1968. A teacher was fired for a letter to the newspaper criticizing his district’s budget handling. The justices reinstated him, holding that absent proof of false statements knowingly or recklessly made by him, a teacher’s exercise of his right to speak on issues of public importance may not furnish the basis for his dismissal from public employment.

According to the opinion, available here, the Court framed the problem as weighing the teacher’s interest in commenting as a citizen on public matters against the state’s interest, as employer, in running an efficient public service. Connick v. Myers later added a threshold: the speech has to touch a matter of public concern, not a purely personal grievance.

Then came Garcetti v. Ceballos in 2006, and it narrowed things. A deputy district attorney, Richard Ceballos, was retaliated against after an internal memo questioning a warrant affidavit. Justice Kennedy, writing for the Court, held that “when public employees make statements pursuant to their official duties, the employees are not speaking as citizens for First Amendment purposes, and the Constitution does not insulate their communications from employer discipline.”

Lane v. Franks confirmed the flip side: sworn testimony about misuse of public funds is citizen speech, subject to Pickering balancing, even when the knowledge came from the job.

So walk it through before you post. If you run your agency’s official account and post policy statements, that is official-duties speech and your employer can discipline you freely. If you post from a personal account on your own time about local budget or policing policy, courts are more likely to treat it as citizen speech on a public concern. Your employer then has to show real disruption to justify punishment.

The complication is that social media blurs the two. A post can be personal and still read as reflecting on the agency, especially if your profile names your employer. Racist or harassing posts can trigger workplace-conduct rules on top of the speech analysis. Because most of these cases are decided in lower courts without Supreme Court review, there is no single clean rule, only the framework: are you speaking as an employee or a citizen, is it a matter of public concern, and can the employer prove harm.

Where Your Rights Are Strongest, and What Is Coming Next

For most people, the clearest protection is the oldest one: showing up in public to say something. The ACLU puts it simply, noting that rights are strongest in what are known as “traditional public forums,” such as streets, sidewalks, and parks.

In those spaces, the government has only a limited number of grounds on which it can step in. Those are rules on when, where, and how you protest that apply to everyone regardless of message, and actual violence, which is in no way protected. Peaceful criticism of the government sits at the center of what the amendment secures.

The harder frontier is technology the doctrine never anticipated. The National Conference of State Legislatures reports that more than two dozen states have enacted statutes addressing political deepfakes, often banning distribution near elections or requiring disclosures. One California statute was already struck down for prohibiting expression likely to hurt a candidate’s chances, a reminder that a new harm still gets judged by old tests.

That is the unresolved tension. The tests were built for leaflets and broadcast ads, and now they have to govern algorithmic feeds and AI-generated video.

Counterman’s recklessness standard is being fought over in stalking cases. Murthy left the jawboning question open for the next plaintiff who can prove coercion and standing. The deepfake statutes are heading toward challenges that will test how far disclosure rules can reach.

The framework you started with holds up through all of it. Ask who is trying to control the speech, what kind it is, and how the restriction works. The answers change as the technology does. The questions do not.

Our articles make government information more accessible. Please consult a qualified professional for financial, legal, or health advice specific to your circumstances.