Supreme Court Upholds TikTok Ban, and Domestic and Foreign Censors Rejoice–TikTok v. Garland

In 2024, Congress enacted, and President Biden signed, the Protecting Americans from Foreign Adversary Controlled Applications Act, which bans “foreign adversary” ownership of certain types of Internet services. The bill specifically bans Bytedance/TikTok by name. TikTok and its users challenged the law as applied to them, but the law specified an unusual process for judicial review of its legitimacy that sped up and compressed the normal review process.

In December, the DC Circuit unanimously upheld the ban. Six weeks later, the Supreme Court–moving at lightning speed by its standards–unanimously affirmed the ban in a per curiam opinion. TikTok Inc. v. Garland, 604 U. S. __ (2025).

This ruling might be only a short-term blip for TikTok, but it’s seismic for the Internet.

With respect to TikTok, both President Biden and incoming President Trump have since reconsidered their anti-TikTok position. (Trump 1.0 banned TikTok via an EO; President Biden signed this bill). TikTok seems likely to get a free pass from the Biden administration for the one day before the transfer of power. If so, ban enforcement will kick over to President Trump, who is famously unpredictable but sees personal benefits to disregarding the law. (Not coincidentally, the TikTok CEO is getting a premium seat at Trump 2.0’s inauguration). So will TikTok be banned in the US? Perhaps in the short run by the app stores, but otherwise, TikTok now has two presidents working on its behalf to undo the censorship imposed by Congress and the Supreme Court.

With respect to the Internet, the Supreme Court’s ruling will foster more online censorship. The Supreme Court endorses a “Fortress USA” approach, i.e., legislatures can impose special speech rules that create a bordered Internet customized for the US and more censored than the Internet in other countries. Many foreign governments are already pursuing bordered and heavily censorial rules for their country. Now, those governments can point to Congress’ and the Supreme Court’s endorsement of the practice as validation of the tactic’s legitimacy.

The Supreme Court also endorses the deputization of First Amendment-protected actors (app stores) to enforce this Fortress USA censorship. (Note: the app stores sat this litigation out, letting others to fight their battles for them). Legislatures are exploring a variety of ways of turning third-party intermediaries into their censorship proxies, and the Supreme Court green-lights that here (sub silento–app stores aren’t mentioned once in the opinion).

Finally, the Supreme Court’s ruling, though heavily caveated, provides a roadmap for other legislators to pursue future censorship. Some of the tricks those legislatures can try: set a short timeline on constitutional challenges, target the ownership of media enterprises for censorship rather than enacting censorship outright, and pretextually justify the law as advancing “privacy” or speculative “national security” interests. The Supreme Court might walk back some of these potential exceptions in future rulings. Until then, legislatures can and will continue to aggressively target free speech online with the Supreme Court’s encouragement, and we will all be the poorer for it.

Introduction: the Splinternets Problem

The TikTok ban fits into a broader genre of Internet censorship that imposes geographically-based “borders” on a communications network where the technology doesn’t inherently encode those technological borders. Countries are routinely creating unique variations of the Internet for their country that differ from other countries’ Internets due to those country-specific regulations. For example, China’s Internet looks very different than the rest of the world because of heavy government censorship. (Chinese residents use VPNs to try to recreate a non-Chinese Internet, but the government tries to shut down the VPNs to prevent this). The sharding of a single Internet into hundreds of country-specific Internets is called the “splinternets.”

The TikTok ban is a splinternets bill. It seeks to kick TikTok out of the US, even if it’s available in other countries. This makes the U.S. Internet looks different than the Internet elsewhere.

To be clear, the US already has a US-specific Internet. That’s partially attributable to the First Amendment and Section 230, which have enabled online conversations to take place in the US that are foreclosed in other countries due to their more restrictive speech rules. Now, Congress has shifted the speech gears into reverse. The US is taking a more speech-restrictive policy than other countries.

But perhaps not for long. The Supreme Court’s ruling will surely embolden other countries to ban TikTok. While they are at it, the countries will ban other apps too–including US apps. The proliferation of app bans by other countries will surely hurt US interests in the long run far more than TikTok ever could have.

In particular, there’s a painful irony about Congress’ purported rationale justifying the TikTok ban: that the Chinese government could slurp up too much data about US citizens and thereby jeopardize national security. You may recall that the EU repeatedly banned the export of EU residents’ personal data to the US because of concerns that the US government slurps up that data to the detriment of EU residents. If other countries are going to ban apps in an effort to thwart foreign governments from collecting data about their residents, US apps are in serious trouble.

Is the Ban Even a Speech Restriction?

The Supreme Court starts by questioning if a divest-or-be-banned law even restricts speech:

It is not clear that the Act itself directly regulates protected expressive activity, or conduct with an expressive component. Indeed, the Act does not regulate the creator petitioners at all. And it directly regulates ByteDance Ltd. and TikTok Inc. only through the divestiture requirement.

Say what? I can’t believe the speech-restrictive effects of the law is a contested issue (as Justice Sotomayor notes in her concurrence). Nothing regulates expressive activity more dramatically than shutting down a speech venue. That nukes all of the users’ speech occurring in that venue, including all constitutionally protected speech. The court even admits that: “an effective ban on a social media platform with 170 million U. S. users certainly burdens those users’ expressive activity in a non-trivial way.” The ban also nixes the speech venue’s own speech, expressed through their content moderation and editorial policies.

Furthermore, there is no doubt that controlling the identity of media owners affects the speech expressed by that media. We see it all the time when publishers change owners. Think of the policy changes at the Washington Post under Bezos and the Los Angeles Times under Soon-Shiong. The editorial impacts of media ownership is a primary justification for the bans of foreign ownership of broadcasters; and the TikTok ban was intended to prevent the Chinese government from jiggering the algorithms to change their editorial impact. I don’t understand how the court could be so milquetoast about the expressive consequences of picking-and-choosing media owners.

Plus, if controlling media ownership doesn’t affect expressive activity, then legislatures use ownership divestments as a tool for censorship. Imagine a hypothetical law saying that a print publisher must divest ownership unless they censor all [pro-Republican] content? (Feel free to replace the word Republican with whatever you think would be the most noxious angle for censorship). According to this court, this law might not restrict expression because it just dictates ownership of the speech venue. 🙄

To limit that risk, the court says the TikTok ban is different because of “the Act’s focus on a foreign government, the congressionally determined adversary relationship between that foreign government and the United States, and the causal steps between the regulations and the alleged burden on protected speech.” Are those differences meaningful at all? To me, they sound like a roadmap for other legislatures to adopt some pretextual legislative findings and go forth with their censorship.

Standard of Scrutiny

The court says the law should be reviewed using intermediate scrutiny. It summarizes that “the challenged provisions are facially content neutral and are justified by a content-neutral rationale.” The court explains:

Not a Speech Restriction

The court says the law is content-neutral because it’s not based on TikTok’s speech:

They impose TikTok-specific prohibitions due to a foreign adversary’s control over the platform and make divestiture a prerequisite for the platform’s continued operation in the United States. They do not target particular speech based upon its content or regulate speech based on its function or purpose. Nor do they impose a “restriction, penalty, or burden” by reason of content on TikTok—a conclusion confirmed by the fact that petitioners “cannot avoid or mitigate” the effects of the Act by altering their speech.

The court seems to be saying that the government had no proof that the Chinese government jiggered TikTok’s algorithm, so Congress isn’t picking TikTok’s speech. This confused me. I don’t see how permitting interventions based on speculations about possible future content choices is any less pernicious than intervention due to past content choices. They are both content-based!

Privacy Concerns Aren’t Content-Based

The court says that privacy concerns are content-neutral:

preventing China from collecting vast amounts of sensitive data from 170 million U. S. TikTok users…is decidedly content agnostic. It neither references the content of speech on TikTok nor reflects disagreement with the message such speech conveys.

I mean, censors routinely pretextually package their censorship laws under the “privacy” aegis. This opinion encourages more of that.

A less credulous court would have said that the privacy-as-pretext justification is apparent because the divest-or-ban approach is obviously underinclusive as a privacy regime, not well-tailored to advance the privacy goal, and comes with unacceptable levels of collateral damage (all of the constitutionally protected speech that gets nuked from the ban). The court did not do that.

Targeting TikTok By Name

The court doesn’t care that the law expressly regulates TikTok by name:

requiring divestiture for the purpose of preventing a foreign adversary from accessing the sensitive data of 170 million U. S. TikTok users is not “a subtle means of exercising a content preference.” The prohibitions, TikTok-specific designation, and divestiture requirement regulate TikTok based on a content-neutral data collection interest. And TikTok has special characteristics—a foreign adversary’s ability to leverage its control over the platform to collect vast amounts of personal data from 170 million U. S. users—that justify this differential treatment.

So….other legislatures can call out publishers by name, make legislative findings disparaging the publisher (using selective presentation of evidence, as legislatures too often do), and try to knock publishers out of the market due to those named entities’ “special characteristics.”

Recognizing the blank check it’s issuing to censors, the court then suggests that there’s a CCP exception to the First Amendment:

we emphasize the inherent narrowness of our holding. Data collection and analysis is a common practice in this digital age. But TikTok’s scale and susceptibility to foreign adversary control, together with the vast swaths of sensitive data the platform collects, justify differential treatment to address the Government’s national security concerns. A law targeting any other speaker would by necessity entail a distinct inquiry and separate considerations.

Ha ha. I guarantee you that the pro-censorship folks’ eyes will glaze over that paragraph. They will be enthusiastically mining this opinion for the pro-censorship pieces without seeing anything “inherently narrow” about this ruling.

Applying Intermediate Scrutiny

Based on the court’s assessment that this is a content-neutral law, the court applies intermediate scrutiny.

Government Interest

The court says Congress had an important government interest in banning TikTok:

The Act’s prohibitions and divestiture requirement are designed to prevent China—a designated foreign adversary—from leveraging its control over ByteDance Ltd. to capture the personal data of U. S. TikTok users. This objective qualifies as an important Government interest under intermediate scrutiny.

Because of the breadth of that statement, the court enumerates the specific things cited by Congress that the court accepts as threats to the US (the court disregards the secret evidence about TikTok’s nefariousness that has never been shared publicly):

The platform collects extensive personal information from and about its users. See H. R. Rep., at 3 (Public reporting has suggested that TikTok’s “data collection practices extend to age, phone number, precise location, internet address, device used, phone contacts, social network connections, the content of private messages sent through the application, and videos watched.”); 1 App. 241 (Draft National Security Agreement noting that TikTok collects user data, user content, behavioral data (including “keystroke patterns and rhythms”), and device and network data (including device contacts and calendars)). If, for example, a user allows TikTok access to the user’s phone contact list to connect with others on the platform, TikTok can access “any data stored in the user’s contact list,” including names, contact information, contact photos, job titles, and notes. 2 id., at 659. Access to such detailed information about U. S. users, the Government worries, may enable “China to track the locations of Federal employees and contractors, build dossiers of personal information for blackmail, and conduct corporate espionage.” 3 CFR 412. And Chinese law enables China to require companies to surrender data to the government, “making companies headquartered there an espionage tool” of China. H. R. Rep., at 4….

the Government’s TikTok-related data collection concerns do not exist in isolation. The record reflects that China “has engaged in extensive and yearslong efforts to accumulate structured datasets, in particular on U. S. persons, to support its intelligence and counterintelligence operations.”

Doesn’t this evidence generically describe how most UGC services operate? In other words, social media by its nature is a giant data collection engine. The court suggests that those ordinary operations are enough to justify censorship? If the ordinary operation of social media services pose national security risks due to the hypothetical potential of data to reach the bad guys, it seems like legislatures could push all kinds of censorship trying to control the ordinary operation of social media.

Worse, the court accepts the Congressional evidence credulously, even though it’s all speculative concerns about hypothetical future harms. Congress sold a bunch of “maybes” and “what ifs” to the Supreme Court. The Sinophobia worked.

Due to that Sinophobia, the court doesn’t care about the ban’s underinclusiveness at starving the Chinese government of sensitive data about US residents: “the Government had good reason to single out TikTok for special treatment.”

Tailoring

The court is fine with how the ban nominally protects sensitive data of US residents:

The provisions clearly serve the Government’s data collection interest “in a direct and effective way.” The prohibitions account for the fact that, absent a qualified divestiture, TikTok’s very operation in the United States implicates the Government’s data collection concerns, while the requirements that make a divestiture “qualified” ensure that those concerns are addressed before TikTok resumes U. S. operations. Neither the prohibitions nor the divestiture requirement, moreover, is “substantially broader than necessary to achieve” this national security objective. Rather than ban TikTok outright, the Act imposes a conditional ban. The prohibitions prevent China from gathering data from U. S. TikTok users unless and until a qualified divestiture severs China’s control.

It’s jarring that the court says shutting down a speech venue isn’t “broader than necessary” to protect privacy. It reminds of the classic line in First Amendment jurisprudence: “to burn the house to roast the pig.” The court is willing to burn down the TikTok house, and all of the speech it enables, to roast the speculative pig of Chinese government data transfers.

Worse, the court took Congress’ bait that it was a “conditional ban.” In my book, a ban is a ban, but not here. Censorship-minded legislatures will be very excited about imposing “conditional bans” on speech and then doing a wink-wink in the courts, “but our ban wasn’t an UNCONDITIONAL ban, now was it?”

TikTok pointed out that Congress had an array of more effective tools to block the leeching of sensitive data about US governments to foreign adversaries. Indeed, the bill–in a part that everyone has been ignoring–simultaneously created 15 USC 9901, which bans data brokers from transferring sensitive PII to foreign adversaries. This provision proves that Congress had other available tools beyond banning a speech venue, but the court doesn’t care:

Petitioners’ proposed alternatives ignore the “latitude” we afford the Government to design regulatory solutions to address content-neutral interests….The validity of the challenged provisions does not turn on whether we agree with the Government’s conclusion that its chosen regulatory path is best or “most appropriate.”

As the expression goes, “close enough for government work“, I guess.

The Algorithmic Control Justification

The evidence is overwhelming that Congress feared the Chinese government would jigger TikTok’s algorithm. As just one example, the bill sponsor Rep. Mike Gallagher said: “[Data security is] a serious threat, but I actually think the greater concern is the propaganda threat.”

The court is unperturbed:

The record before us adequately supports the conclusion that Congress would have passed the challenged provisions based on the data collection justification alone…nothing in the legislative record suggests that data collection was anything but an overriding congressional concern.

Thus, Congress’ pretextual justification of “privacy” worked, despite demonstrable proof of censorial intent. Future legislatures will take note. They will feel free to blab all they want about your censorship goals, so long as they sprinkle the law with lots of references to “privacy” and “security.”

Court’s Conclusion

The court concludes:

There is no doubt that, for more than 170 million Americans, TikTok offers a distinctive and expansive outlet for expression, means of engagement, and source of community. But Congress has determined that divestiture is necessary to address its well-supported national security concerns regarding TikTok’s data collection practices and relationship with a foreign adversary.

Bye bye TikTok. 🙋‍♂️

Sotomayor Concurrence

Justice Sotomayor said there’s “no doubt” that “the Act implicates the First Amendment.”

Gorsuch Concurrence

Justice Gorsuch raises some concerns about the court’s analysis. He says:

One man’s “covert content manipulation” is another’s “editorial discretion.” Journalists, publishers, and speakers of all kinds routinely make less-than-transparent judgments about what stories to tell and how to tell them. Without question, the First Amendment has much to say about the right to make those choices…

I harbor serious reservations about whether the law before us is “content neutral” and thus escapes “strict scrutiny.”

And yet, he might very well uphold the law on strict scrutiny:

I am persuaded that the law before us seeks to serve a compelling interest: preventing a foreign country, designated by Congress and the President as an adversary of our Nation, from harvesting vast troves of personal information about tens of millions of Americans…

Without doubt, the remedy Congress and the President chose here is dramatic. The law may require TikTok’s parent company to divest or (effectively) shutter its U. S. operations. But before seeking to impose that remedy, the coordinate branches spent years in negotiations with TikTok exploring alternatives and ultimately found them wanting.

at this time and under these constraints, the problem appears real and the response to it not unconstitutional….Speaking with and in favor of a foreign adversary is one thing. Allowing a foreign adversary to spy on Americans is another.

Sinophobia trumps the First Amendment.

Implications

Will This Ruling Affect Other Entities? The law doesn’t only ban TikTok, though TikTok and its family was the only target referenced by name. The president (soon to be Trump) can designate other apps for divest-or-ban through a statutorily specified process. That’s a dangerous weapon to put into Trump’s hands because he can use it as leverage to extract concessions from other apps. In theory, he might use that power to protect our nation’s interests. More likely, he would use that power to benefit himself or his oligarch bros.

If any other app is targeted, the court’s opinion doesn’t necessarily predict that the divest-or-ban mandate would be constitutional. The court only reviewed the law as applied to TikTok, and the legislative findings applicable to TikTok won’t extend to other targeted apps. I’m sure the court viewed this opinion is a one-off burning of TikTok and not burning anyone else. If the law ever gets used for any other purpose, we’ll see if that’s what they actually achieved.

The Rushed Process. Congress set up a rushed process for reviewing this law. First, it dictated that the review would start in the D.C. Circuit, rather than in a district court as is typical. The D.C. Circuit did an OK job acting in the role that is normally occupied by a district court, but that necessarily led to truncated fact-finding and motion practice. Second, the ban goes into effect in 9 months from passage–an extraordinarily short timeline for the law to go through two layers of judicial review (even though Congress shaved off the third layer of review at the district court level).

Because the D.C. Circuit proceedings took most of those 9 months, the Supreme Court only had about 6 weeks from when it received the case until when the ban would go into effect. The Supreme Court could have stayed the law to permit a more orderly review, but instead it tried to do a rush job. Unfortunately, the Supreme Court is not designed for rush jobs, and it doesn’t have much experience doing them. The result is that the advocacy was not as sharp as the Supreme Court is used to getting, and the factual record wasn’t as well vetted as it normally gets. I feel like this process favored upholding the law, especially as the court relied more heavily on Congress’ finding instead of the fact record that would have been developed through normal litigation.

These procedural corner-cuts provide a roadmap for future Congressional efforts to shelter censorial laws from full judicial scrutiny. I wonder if state legislatures might explore this tactic as well (if they can).

Alternative Venues for TikTok Content Creators? The Supreme Court gave mere lip service to how many speakers will lose their ability to engage in constitutionally protected speech due to the ban. In some circumstances, the First Amendment requires the court to consider if there are alternative means of expression, something the court didn’t do here.

There are certainly other venues for TikTok creators. A number of services offer competing apps, but those are imperfect substitutes. First, every UGC site has unique design features and affordances that shape the content they create. Even if it’s the exact same creator trying to reach the same type of audience, content from one venue doesn’t always translate well to other venues. Second, TikTok’s “for you” algorithm is a unique way of helping users discover speech. That creates audiences that the same content creator could not reach on other services, and that can translate into important financial incentives. The algorithm also reflects TikTok’s unique editorial expression that differs from the editorial expression of other services’ algorithms.

The net effect is that a TikTok ban has significant collateral damage. It shuts down constitutionally protected conversations that won’t migrate elsewhere; that content will be lost forever.

The TikTok Diaspora. In anticipation of the ban, some TikTok users have migrated elsewhere. TikTok urged its users to migrate to Lemon8, but Lemon8 should be subject to same ban as TikTok. Other TikTok users migrated to Rednote, which is ironic because Rednote is more obviously connected to the Chinese government than TikTok is. In other words, Congress drove US TikTok users from a possible national security threat (TikTok) to a more obvious national security threat (Rednote). Nice work, Congress.

Picking Winners and Losers in the Speech Marketplace. Congress’ power flex is deeply chilling to First Amendment interests. Essentially, Congress is kicking TikTok out of the speech marketplace because it doesn’t like who owns TikTok. We’ve tolerated Congress’ control over foreign ownership of broadcasters because of the limited number of broadcast licenses, but even that restriction gives the government troubling control over who get the right to broadcast. When there isn’t a scheme of restricted licenses, giving Congress (or any legislatures) the power to pick-and-choose who owns the printing presses is extraordinarily censorial. It’s ripe for partisan abuse, too.

Democrats’ Own-Goals. The Democrats committed a series of own-goals in enacting this law. The Democrats could have scuttled this bill if they wanted, but didn’t.

As a result, the Democrats fully put their censorship tendencies on display for voters. (Censorship is bipartisan, so this isn’t a new thing). They also sent a terrible message to young voters that they wanted to rip TikTok out of their hands. When Trump flipflopped on TikTok (after unsuccessfully banning it during his 1.0 administration), it enabled Trump to suggest that he was more in tune with the “youts” than Kamala. It also helped him repeatedly claim to be a free speech advocate–which he absolutely is not, but there was a kernel of truth to it in this instance. I doubt this issue cost the presidential election, but the Democrats certainly did themselves no favors.

The even worse own-goal comes from President Biden’s openness to deferring enforcement of the ban until after he is out of office:

• Biden enthusiastically signed the law when it passed, so flipping on this issue makes him look indecisive and weak.
• Biden signed relatively few major bills during the last two years of his term, so he undercut one of those accomplishments by flipping. Biden has a lot of strong features to claim in his historical legacy, but this is not one of them.
• Biden undermined the DOJ’s advocacy about the national security risks. The DOJ made the sale in court, and Biden’s flip-flop leaves them twisting in the wind that the fears were overstated.
• Biden positions Trump to come in as the free speech savior, making Trump look smarter than he actually is and making the Democrats look dumb.
• The flip-flop undermines our government. Enacting a law on national security grounds, and then saying “just kidding” after the Supreme Court takes the bait, is highly corrosive on our ability to trust our government.

Great job, Democrats.

The App Stores’ Acquiescence. The app stores no-showed in this litigation. They would have surely preferred the ban to be struck down, but they will quietly and without protest implement the ban (unless they get sufficient DOJ assurances that they face no legal risk).

It’s a sign of how easily democratic principles acquiesce to abusive government practices. The app stores don’t really care what censorial rules apply to them and what apps they are allowed to carry so long as they are allowed to remain in business. Traditionally, bookstores have been vanguards in free speech litigation, carrying the First Amendment banner in the face of censorship. App stores? Not so much.