The TikTok Case Will Be Determined by What’s Behind the Government’s Black Lines

Late last month, the U.S. government filed its brief in the litigation challenging the Protecting Americans from Foreign Adversary Controlled Applications Act (PAFACAA), which would ban TikTok in the United States unless ByteDance, its parent company, divests by early 2025. Overall, the brief is compelling, making many of the strongest arguments for the law’s constitutionality, including the law’s divestment option, the historical precedent for restricting foreign ownership of media platforms, and the importance of deferring to political branches’ risk calculations.

But the government’s central argument is the national security case: TikTok’s partial control by ByteDance, and thus potentially by the Chinese government, simply poses too great a risk to Americans’ data privacy and information integrity. This argument relies heavily on classified information that the government has redacted, making the evidence available only to the court itself and not to TikTok or, just as important, the public. Thus, even if the government ultimately succeeds in defending the law (as we think it likely will, though others may of course reasonably take a different view), it risks doing so in a way that will satisfy few observers, including those sympathetic to its position.

The composition of the government’s legal team—both in number and variety—indicates just how critical the national security evidence is to the government’s case. The Department of Justice is, of course, well represented, with lawyers from the Appellate Staff and the Federal Programs Branch of the Civil Division joined by the National Security Division, the FBI, and the Office of the Director of National Intelligence (ODNI). An appendix accompanies the brief, containing three declarations—one from ODNI, one from the FBI, and the third from Justice—all detailing the national security threat that TikTok poses.

From the unredacted portions of the brief, readers are able to get a sense of the broad strokes of the government’s national security argument. The government asserts that China aims to become a preeminent global power by undercutting U.S. influence and fostering authoritarian norms. Within this context, the government portrays TikTok as a potential tool for Chinese espionage and influence operations, highlighting TikTok’s extensive data collection practices, its ability to manipulate its recommendation algorithm, and instances of censorship at the behest of the Chinese government. The government contends that these factors create significant risks to U.S. national security.

But in this section of the government’s brief, nearly half the content is redacted. Sometimes multiple pages in a row are blacked out. And one critical subsection, discussing China’s incentives to incorporate TikTok into its offensive cyber strategy, is almost entirely classified, except for three short paragraphs.

The evidence that the government did release, which primarily consists of public statements from intelligence officials and newspaper reporting, is somewhat vague and generic. Without access to the redacted evidence, many of the government’s claims—for example, that China is well positioned to influence the U.S. component of TikTok—read as unsubstantiated (though hardly implausible) assertions about China’s strategic goals, ByteDance’s potential ties to the Chinese government, and TikTok’s data collection practices.

Because the government redacted entire paragraphs of evidence, it not only obscured specific classified details but also often made it difficult to determine, even at the most abstract level, what the classified evidence was even attempting to establish in the first place. As a result, it is challenging for both the TikTok legal team and outside observers to fully grasp or critically evaluate the government’s case.

In a separate motion, the government justifies its redactions on two main grounds. First, it argues that federal law prohibits the public disclosure of information classified as “secret” or “top secret.” Second, it contends that the U.S. Court of Appeals for the D.C. Circuit has repeatedly allowed the government to file redacted classified material in the past, and it points to many different cases where the court rejected challenges to its redaction of evidence.

TikTok filed a blistering response seeking to block the government from relying on secret evidence. Opening with Justice Louis Brandeis’s famous observation that “[s]unlight is said to be the best of disinfectants,” it accuses the government of wanting “this case to be litigated in the dark—with the free speech rights of Petitioners and 170 million Americans hanging in the balance.”

Some of TikTok’s arguments are unconvincing. For instance, it argues that the court should reject the government’s classified evidence because it was not part of the formal legislative record before Congress enacted PAFACAA. However, this argument overlooks the many classified national security and intelligence briefings that Congress received and that undoubtedly played a major role in the legislation’s passage. Even if the government is now bolstering its case with specific pieces of evidence that were not available to Congress, the court is unlikely to strike down a high-profile piece of national security legislation for that reason.

But other parts of TikTok’s argument are more persuasive. For example, TikTok points to what it characterizes as multiple factual errors in the unclassified portion of the government’s brief, arguing that these undermine the credibility of the classified portions. Inaccurate redacted information would be particularly concerning, because those claims can’t be tested in the adversarial legal process. Specifically, TikTok disputes the government’s characterization that TikTok is owned by a Chinese company, noting that ByteDance Ltd., TikTok’s ultimate owner, is a “Cayman Islands–incorporated holding company majority-owned by global institutional investors.” TikTok also challenges the government’s assertions that it collects users’ precise location data and that its recommendation engine is located in China (TikTok argues that it’s actually in an Oracle-operated cloud).

The issue is not whether TikTok or the government is ultimately correct on the specific factual points. Rather, these factual questions, which lie at the center of the case, are difficult for the D.C. Circuit to evaluate because the government relies so heavily on secret evidence—especially in the absence of a trial court factual record, since PAFACAA requires challenges to be brought directly in the D.C. Circuit.

More fundamentally, TikTok is right that, as a general matter, secret evidence is highly disfavored. In addition to the evidentiary issues discussed above, secret evidence undermines public confidence in the legal system, because the public doesn’t know on what basis the government is acting. This is particularly concerning in a case that could impact every one of the 170 million Americans who use TikTok, including many who rely on it for their livelihoods.

Furthermore, if the court’s decision is based on secret evidence, it will be difficult to apply the ruling to future cases, as the factual details driving the outcome will remain hidden. If the case reaches the Supreme Court—which it almost certainly will—it will be one of the most important First Amendment cases in recent memory. But the reliance on secret evidence could significantly weaken its value as precedent. Without access to the underlying facts, future courts will struggle to interpret or apply the ruling in similar cases. And given that TikTok is hardly the only foreign-controlled platform that could potentially pose national security threats, this issue is not disappearing any time soon.

TikTok is also correct that secret evidence is dangerous when fundamental rights like free speech are involved. The government argues that PAFACAA has “at most … an incidental effect” on TikTok users’ First Amendment rights, because the law targets only TikTok itself, not any specific content, and thus “any preference these petitioners may have for using TikTok over those other platforms does not create a constitutional right to TikTok.”

However, banning TikTok is not analogous to a time, place, and manner restriction telling protesters which street corner they can yell from. TikTok has its own unique features—its algorithm, community, and design—that make it not interchangeable in the way that, for example, a cell phone or internet provider might be. That of course does not mean that TikTok users’ preference for the platform is inviolate—the government is right that other platforms provide roughly similar experiences. But it’s a legitimate First Amendment interest that should be taken seriously in the course of litigation.

In ordinary circumstances, notwithstanding the legitimate interests in public evidence, the government’s national security argument would likely prevail. As one of us has written previously:

Certainly the D.C. Circuit can consider classified evidence if it wants to—it routinely did so in the course of its many Guantananamo habeas cases. More recently, the Ninth Circuit reviewed classified information as part of its decision rejecting Twitter’s request for the ability to publish more detailed transparency reports on government surveillance orders. The Ninth Circuit noted, “While we are not at liberty to disclose the contents of the classified materials that we reviewed, our analysis under the narrow tailoring prong [of the First Amendment] depends principally on the knowledge we gleaned from our review of that material.” This suggests an answer to a question that Jameel Jaffer, executive director of the Knight First Amendment Institute at Columbia University, rhetorically posed in a Twitter exchange with me over the pending TikTok litigation: “I mean, what’s the opinion going to say? ‘Trust us, your First Amendment rights haven't been violated?’ I think the government will have to justify the law on the public record.” Some version of “trust us” may well be exactly what the D.C. Circuit says.

Recognizing that it is fighting an uphill battle, TikTok has suggested a middle ground alternative, recommending that the court appoint a “special master who, after hearing from the parties, would recommend to this Court what procedural protections—if any—could mitigate the prejudice to Petitioners from admitting ex parte the government’s classified submissions.”

At first glance, this might seem like a Hail Mary play. But because PAFACAA is such a high-profile law, with obviously huge First Amendment implications, the court may seriously consider it.

Of course, as TikTok recognizes, the appointment of a special master would considerably delay the litigation and likely require the court to enjoin PAFACAA in the meantime. It would also raise ancillary questions, such as whether the injunction would freeze or reset the divestment clock, and the court might be hesitant to add a host of procedural complications to an already difficult constitutional case. Moreover, even the appointment of a special master would not completely resolve the transparency concerns, because many of the details would likely remain classified, even after another level of judicial review.

But this might be the kind of case where these costs are worth paying. The government’s emphasis on national security is absolutely correct, not because the First Amendment issues aren’t important, but precisely because they are, and therefore should be overridden only by national security concerns of the highest order. It would not be unreasonable for the court to want to convince itself—and credibly signal to the challengers and the public—that it has done everything it could to check the government’s work, even if it means that a substantively thorny case is sidetracked by a procedural fight over secret evidence.