TikTok v. Garland and the Hard Clash of Free Speech and National Security

Most hard cases are hard because the law is unclear. This one was hard because the law was clear and the values collided. A free-speech conservative and a China hawk, often the same person, found themselves arguing with themselves, and both sides had good lines. It is not often that a dance-video app produces a constitutional seminar, but here we are.

I am going to treat that honestly. The case is TikTok Inc. v. Garland, decided January 17, 2025, and it asked whether Congress may require a company owned by a firm in a country that is America's chief strategic rival to sell its American operation or leave the market, when tens of millions of Americans use the app to speak and to listen. The Court said yes, unanimously. I think the result is defensible. I think the reasoning has limits that matter more than the result. And I think conservatives on both sides had arguments that deserve to be stated at full strength.

The facts and the statute

TikTok is a short-video platform operated in the United States by TikTok Inc. Its ultimate parent is ByteDance Ltd., a company founded in China. The app was enormously popular with Americans. The concern, shared by lawmakers in both parties across several years, was that a company subject to Chinese law could be made to hand over data about American users, or to tilt what Americans see, at the direction of the government in Beijing.

Congress responded with the Protecting Americans from Foreign Adversary Controlled Applications Act, enacted in April 2024 as part of a larger national security package and signed by President Biden. The statute does not use the word "ban" in its operative sections. It makes it unlawful for app stores and internet hosting services to distribute, maintain or update a "foreign adversary controlled application," which the law defines to include apps operated by TikTok and ByteDance and, by a designation process, other apps controlled by a foreign adversary and found to pose a national security threat. The law gave a window of 270 days, ending on January 19, 2025, for a "qualified divestiture," a sale that severed the app from foreign adversary control, with a possible extension of up to 90 days if the President certified that progress toward a sale was being made.

TikTok, ByteDance, and a group of American users and creators challenged the law directly in the D.C. Circuit, which has original jurisdiction under the statute's review provision. The D.C. Circuit upheld it in December 2024. The Supreme Court heard argument on January 10, 2025, a week before the deadline, and decided a week later.

What the Court held

The opinion was per curiam, meaning issued in the Court's name rather than by one Justice, and unanimous.

The Court did not announce that the First Amendment was irrelevant. It assumed, without deciding, that the law is subject to First Amendment scrutiny at all. It found the provisions aimed at TikTok content neutral as applied to the challengers, rejected strict scrutiny, held that no more than intermediate scrutiny was in order, and concluded that the statute satisfies it. (The D.C. Circuit below had assumed strict scrutiny and upheld the law anyway.)

The Court relied chiefly on the interest in preventing a foreign adversary, here China, from collecting data on tens of millions of Americans, and found that interest alone sufficient to support Congress's action. It did not decide whether the government's second argument, that Chinese control could be used to covertly manipulate the content Americans see, is content neutral, and it did not rest on it. That is a significant limit. The Court's reasoning leaned on the data concern and on the foreign-adversary setting, not on any concern about what viewpoints the app's algorithm might favour. The Court also stressed that its decision addressed this particular law and this particular case, and cautioned against reading it as a broad statement about how the First Amendment applies to new technology.

Justice Sotomayor and Justice Gorsuch each wrote separately. Sotomayor, concurring in part and in the judgment, would not have assumed that the First Amendment is implicated: she thought precedent settles that TikTok's curation of content is expressive activity and that the law burdens it, and she still agreed the law survives. Gorsuch, concurring in the judgment, called his observations tentative because of the expedited schedule, expressed serious reservations about whether the law is content neutral, and said the Court rightly refrained from endorsing the covert-manipulation interest. He did find a compelling interest in stopping a foreign adversary from harvesting Americans' data, and found the law appears appropriately tailored. If you rely on that account for anything that matters, read both opinions at the Supreme Court's website, because a one-sentence summary of a concurrence is a risky thing.

Premises worth getting straight

Let me put the argument as a set of numbered premises, because it is the cleanest way to see what is and is not at stake.

1. The First Amendment restrains the government's power over speech, including the speech of platforms and of listeners. The right to receive information is part of the freedom of speech. In Lamont v. Postmaster General (1965), the Court struck down a statute that required the Post Office to detain "communist political propaganda" from abroad unless the addressee asked for it. The Court's reasoning, that the burden on recipients was itself a First Amendment problem, has been cited for decades.

1. But the Constitution is not a suicide pact on matters of foreign threat, and Congress has real power over foreign commerce and national defence. Article I gives Congress the power to regulate commerce with foreign nations, and the national security powers of the political branches are deep. Courts defer, within limits, on matters involving foreign adversaries. In Holder v. Humanitarian Law Project (2010) the Court upheld a ban on providing material support to designated foreign terrorist organisations even as applied to speech-like assistance, on the strength of deference to the executive's judgment about foreign threats.

1. The relevant question is the target of the law. A law aimed at what is said is the paradigm of unconstitutional regulation. A law aimed at who controls the speaker, specifically a foreign adversary, is a different object. Federal law has long limited foreign ownership of broadcast licences, for example, and the Court has not treated those limits as censorship.

1. The result depends on tailoring. A law that forces divestment of foreign control leaves the speech intact, because the platform, its users and the content can all continue under American ownership. A law that simply barred the content would be a different matter.

Those four premises produce the result the Court reached, and I think they explain why the Justices all agreed. They do not, however, settle every conservative worry.

The conservative case for the law

Here is the argument at its strongest, as its defenders would state it.

The Chinese Communist Party is not an ordinary government. It runs a one-party state that censors its own people, imprisons dissenters, and treats private companies as instruments of state power. China's National Intelligence Law of 2017 requires organisations and citizens to support and cooperate with national intelligence work, and Article 7 says so in terms. A company headquartered there, or whose parent is, cannot guarantee that it will refuse a demand from the state. We have a separate look at how that state treats its own citizens' beliefs in How China's Communist Party licenses churches and jails pastors.

Now consider what an app on a hundred million phones can do. It can collect location data, contacts, device information and behavioural data. It can promote and suppress content through an algorithm that no outsider can audit. A rival power does not need to censor American speech outright. It needs only to nudge what millions of people see, at moments that matter. The risk is not that the app will be used that way. It is that the capability sits in the hands of a government hostile to American interests, and that a prudent country does not leave a capability like that in a rival's pocket.

The defenders add a point of fairness. China bars American social platforms from operating within its borders. The asymmetry, one-way openness, is not an argument that America should copy Beijing's censorship. It is an argument that reciprocity is a legitimate consideration in the treatment of foreign-controlled media. And the remedy chosen was the mildest one on offer: sell, and keep operating.

Finally, they point to precedent that conservatives like: restrictions on foreign ownership of broadcast stations, on foreign contributions to American campaigns, and on foreign agents operating without registration. A country is allowed to insist that the channels of its own public discourse be owned by people who are not agents of a rival state.

The conservative case against the law

The other side is equally serious, and a good many conservatives made it.

First, the speech. A law that makes it unlawful to distribute an app that tens of millions of Americans use to talk to each other is a very large burden on expression, and conservatives have spent decades insisting that the First Amendment protects the speech of people we dislike. Our own consistency is tested when the speaker is foreign-owned. A rule is only a principle if it protects what we would rather not protect. Free speech that covers only the speakers we like is just a fan club with a constitution.

Second, the precedent. The most dangerous kind of power is the kind that can be reused. A law passed against a Chinese app is a template. Tomorrow's administration could use the designation process, which lets the President identify other "foreign adversary controlled" applications, against a platform it dislikes for other reasons. Conservatives, who have good reason to doubt that Washington's judgments about speech are always disinterested, should be slow to hand it a new category. We know from the platform debates that government pressure on private speech is easy to start and hard to stop, a point made in NRA v. Vullo and the Unanimous Rule Against Government Jawboning.

Third, the evidence. The public record about what the app actually did with Americans' data was thinner than the confidence of the rhetoric. The Court relied on the risk and on the government's judgment, not on a proven abuse. A rule that restricts speech on the strength of an unproven risk, because the risk is serious, is a rule with a very wide range of future application.

Fourth, the alternatives. If the problem is data, there are narrower tools: limits on what data any app may collect and sell, restrictions on data brokers, rules about data storage and access. A national data-privacy law would have addressed the problem for every foreign-owned and domestic app alike, without singling out one platform. Congress chose the targeted remedy, and the targeted remedy is the one that sits closest to the First Amendment line. Beijing is not the only entity that harvests American data; a rule that fits only one company looks more like a response to a headline than a principle.

Fifth, the market. A free-market conservative has to ask why Washington should decide who owns a private company's American arm, and whether forced sales at the end of a deadline set by statute give the buyer all the leverage. There is a reason the party of property rights is uneasy about it.

Where I think the line should sit

I want to commit to a view, because a post that ends with "both sides have a point" is a post that did not do its job.

I think the Court got the result right and the principle partially right. Here is the line I would draw, in three parts.

One: control, not content. A government may restrict the ownership and control of a communications channel by a foreign adversary. A government may not restrict what is said on the channel because it dislikes the message. The Court's decision to rest on data and foreign control, and to decline to rest on content manipulation, honours that line. I would keep it, and I would make Congress meet it every time.

Two: the adversary must be named by statute, with a record and with review. The TikTok law identified the company by name in the statute itself, and Congress built a committee record and left the courts a role. A designation process that lets an executive point at a platform and invoke the law on a thin record is much more dangerous, and it should be narrow, public and reviewable. The more the process looks like discretion, the less it looks like law.

Three: use the least speech-restrictive means that works, and build the record. If the real problem is data, legislate data. If the real problem is covert influence, require disclosure of foreign control and algorithmic transparency. A divest-or-leave rule is a heavy instrument, and it should be saved for a case in which the threat comes from the adversary itself and lesser tools would not do.

By that test, the TikTok law passes, narrowly and for the right reason, which is that the owner is a foreign adversary's reach and not that the content is disliked. A law that did not pass would be one that said "this platform spreads views we do not like."

The historical anchor

The Founders knew foreign influence was a real danger. Washington's Farewell Address of 1796 warned against the "insidious wiles of foreign influence," and said that history and experience prove that foreign influence is "one of the most baneful foes of republican government." Hamilton, in Federalist No. 68, wrote that the Electoral College was designed in part to guard against foreign powers gaining an improper ascendancy in the choice of the President. The Constitution's own Foreign Emoluments Clause, in Article I, section 9, forbids federal officeholders from taking gifts or titles from foreign states without the consent of Congress. The Founders were not naïve about foreign power, and they were not afraid to write it into the text.

They were, however, equally clear that the cure for foreign influence must not be a government licence over American speech. The same generation that wrote Washington's warning also saw what happened to the Alien and Sedition Acts of 1798, which used a foreign-threat rationale to jail American editors. See The Sedition Act of 1798 Jailed a Congressman Who Won Reelection From His Cell. The way to hold both lessons is the one I have described: regulate control by a foreign adversary, never the message.

If you want to see how another case of China and American culture tangled, The NBA and China in 2019 and the Week a Tweet Cost a League Its Voice and Made for Beijing, How the Chinese Market Changed Hollywood Scripts show how a foreign party's market power shaped speech without any American statute at all.

The case in brief

1. In TikTok Inc. v. Garland (January 17, 2025), the Supreme Court unanimously upheld a 2024 statute requiring divestment of TikTok from foreign adversary control, applying no more than intermediate scrutiny and finding the data-collection interest sufficient.
2. The Court rested on data collection by a foreign adversary, and it did not endorse the claim that the government may restrict an app because of what it might show.
3. The conservative case for the law rests on the nature of the Chinese state, the capability a rival would hold, reciprocity, and the long tradition of limiting foreign control of channels of communication.
4. The conservative case against rests on the speech burden, the reusability of the precedent, a thin evidentiary record and the existence of narrower data-privacy tools.
5. The line should sit at foreign control and not content, with an adversary named by statute and subject to review, and the least restrictive means that works.

What you can do: Read the First Amendment and Article I, section 8 together in the Constitution, then ask your representatives, through my representatives, to support a national data-privacy standard that applies to every app instead of a single company.