I. Introduction

Transformative ideas and transformative communicative practices from around the globe have enriched American public discourse and updated how Americans interact with one another. Social media platforms are only the latest means of communicating and engaging. However, foreign-owned social media platforms that offer novel, participative ways for people to engage, communicate, and associate are a form of communications infrastructure that is not adequately protected under US law. Although there is a focus on platforms’ communicative features, there is very little discussion of their associational features.

The Supreme Court has recognized Americans’ right to access foreign ideas. This right is sufficient to ensure access to mass-media forms such as print and broadcasting that depend largely on one-way transmission to audiences. But it is insufficient to regulate participatory social media platforms. Together with mass media, but increasingly dominating it, social media forms the underlying infrastructure of communication in American society. Accessing this infrastructure is distinct from, although connected to, the right to receive ideas. The right to communicative apps as infrastructure is the right to access new ways of communicating and associating. Legal discourse needs to build on the right to access foreign ideas, extending it to the right to access foreign communicative infrastructure.

My goal in this essay is to reframe the value of foreign social media platforms (and potential future AI platforms) in the context of TikTok Inc. v. Garland as unique and creative associational and speech infrastructures and show why associational and speech infrastructures should be treated differently from other foreign infrastructure. The justices adopted an outdated view of the significance of foreign social media—suggesting that a platform’s users are the only people affected by the platform—which led them to underestimate the burden on speech from restricting the app or compelling a change in its ownership. Since lawmakers and experts, including TikTok’s advocates, share this outdated view, the TikTok case is an excellent opportunity to discuss foreign platforms’ role in the American public sphere.

My argument for access to foreign communicative infrastructure extends to all significant foreign communicative apps, including chatbot-centric apps. Political, economic, and cultural influences shape communicative apps, and each app amplifies different kinds of content and offers different ways to engage. Changing foreign apps’ ownership removes these influences and burdens speech in ways that judges should take into account. Some of these questions were discussed in the context of the ‘TikTok ban’ in 2024 and 2025. Although the TikTok question has been settled, TikTok Inc. v. Garland is worth discussing because it presents the Supreme Court’s view of the role of foreign communicative apps in the American public sphere.  Unfortunately, the judgment's implications for social media regulation have been eclipsed by the events that followed it. The justices’ application of the intermediate scrutiny test was based on a dangerously flawed understanding of social media. I focus on the meaning and consequences of the TikTok ruling for the American public sphere, not the actual TikTok divestiture.

II. Communicative Apps as Infrastructure

Infrastructure such as language, script, the press, and broadcasting make human communicative activity possible. Such infrastructure increases shared knowledge and shapes public opinion. Online communicative platforms extend this ability by enabling Americans to share, modify, and consume content in conversation with each other and people around the world. In so doing, the platforms facilitate public discourse, including discourse on other platforms and in the media.

As “shared means to many ends,” most communicative apps are infrastructure. Although other legal scholars have made this argument, more work is necessary on what this means for the sectors these apps serve as infrastructure. In this essay, I discuss the implications for speech of treating significant social media platforms as communicative infrastructure. I use the term “communicative infrastructure” to refer to both the media ecosystem as a whole and its individual parts (e.g., The New York Times, the BBC, TikTok, Google Search, and Google’s AI Overview). However, my focus is on the digital public sphere, which consists of layered infrastructures that are regulated differently, including telecommunications and the open internet.

Legal scholarship on social media’s infrastructural features tends to focus on how platforms are shaped by the market, political institutions, legal frameworks, and public policy, and how they use the data they collect. It facilitates a granular discussion of platforms as infrastructure, and specifically of how platforms’ infrastructural role enables public discourse by affecting which people communicate with each other and how they communicate. I aim to develop the speech-enabling aspect of online communicative platforms’ infrastructural role. Treating communicative apps as infrastructure implies recognizing that these apps have downstream effects beyond the platform. My goal is to contribute to First Amendment scholarship while integrating the valuable insights that scholars of law and political economy offer.

Communicative apps are “structured arrangement[s] that facilitate human [communicative] activity across space and time.” In addition to focusing on the influences on and design of these arrangements, and their harmful downstream effects, it is worth focusing on the extent to which they enable democratic goals. Communicative infrastructure, as well as the absence of state censorship, is necessary for free speech, including public opinion formation and debate. Many democratic constitutions restrict the regulation of the press because the press is a type of communications infrastructure, and such infrastructure supports public discourse.

Widely used communicative products, such as Facebook and TikTok, have additional infrastructural characteristics. They are embedded in other social arrangements and technologies, they invisibly support tasks, they reach beyond single events and practices, new participants acquire familiarity with them by using them, they have a standardizing influence on other infrastructures and tools, and they become visible upon breakdown. These elements are also increasingly found in chatbots that rely on large language models and AI-based creation tools embedded in social media apps.

III. Infrastructure is Different from Ideas

As Jack Balkin has pointed out, the First Amendment prevents state interference with speech, but infrastructure is necessary to ensure that individuals have the freedom and ability to participate in public discourse. Consider the infrastructure required for public discourse: to access materials and inspiration, to consult others, and to circulate ideas and conclusions. Communicative infrastructure—extending from the press to generative AI—is critical to maintaining public discourse through its structural role in disseminating information. This is different from and at least as important as the information itself. Along with the right to access ideas, one needs the means to access ideas.

Furthermore, communicative infrastructure exerts an influence beyond its own systems. When the state restricts the speech of, say, a significant university or the operations of a major social media platform, more people are affected than those directly involved with the university or platform. Universities and platforms, in different ways and for different reasons, make certain associations and speech salient. Major social media platforms influence what stories the traditional media carry, how politicians campaign, and even how democratic institutions share information with the public. Other platforms change their practices to compete.

The infrastructural lens illuminates important elements of the relationship between speech, association, and communicative online apps, including social media platforms and AI-powered chatbots. Communicative online apps play a dual role, enabling communicative relationships at an unprecedented scale while also influencing and shaping them in ways that may impede democracy. For example, social media platforms can change what content is available, ranging from single items of content to individual accounts and even whole feeds, and can use removal, amplification, downranking, and shadow banning to shape the kinds of content that are available. Through repeated nudges, platforms also influence who is put into conversation with whom.

A single entity can play a greater infrastructural role if it dominates the market, as Google does. When Google’s products are unavailable, the public notices. That is a feature of infrastructure—it is seamless and invisible when it works and visible when it stops working. Many people rely on Google. Courts are increasingly finding that this reliance gives Google too much infrastructural power. This is important because not all forms of communicative infrastructure may serve democratic values. Non-state actors can also erode infrastructure, which is why structural regulation, such as anti-trust laws, is necessary to prevent private capture of communicative infrastructure that would stunt the diversity of opinion and group formation.

Social media has a powerful effect on public discourse because it steers communicative relationships. Recognizing this associational power and interpreting the First Amendment to account for it is critical for a healthy public sphere. Up to and including the era of broadcast media, regulating media to maximize its benefits to democracy was seen as consistent with First Amendment values, but the standards changed before the social media cases reached the Supreme Court. Veering away from the idea that communicative infrastructure can be regulated to enable and protect First Amendment interests has left policymakers and judges without theories of how to treat foreign communicative infrastructure. Specifically, treating ideas as medium-agnostic and treating media as ownership-agnostic ignores how far ownership affects the manner in which infrastructures shape speech and association. If foreign ideas are valuable, so are foreign ways of relating and communicating.

IV. The Special Role of Foreign Communicative Infrastructure

Apps shape and are shaped by the communities they engage with, the markets they compete in, the organizations that build them, their access to resources, and local laws, among other things. Tellingly, WhatsApp was co-founded by a man whose Soviet origins taught him the importance of encrypted communication. The Chinese company DeepSeek’s large language models were shaped by resource constraints; the company’s engineers innovated in response to limited access to the powerful new generation of Nvidia chips that power generative AI. Local dynamics in China trigger innovation in different directions. This produced ByteDance’s platform Douyin’s evolving business practices and addictive algorithm for sharing and watching short videos. China has a highly competitive short-video market. ByteDance’s platform for international users, TikTok, has successfully tailored learnings from Douyin for international markets.

As the spread of the short-video format and TikTok’s practices to engage users to other popular communicative platforms shows, foreign innovation can revitalize the infrastructural practices of American communicative apps. Even American social media users who have never engaged with any TikTok videos are affected by changes that American platforms make as they innovate to keep up with ByteDance. The Chinese company is a major competitive presence in a highly concentrated market and triggers changes in the social media ecosystem that can benefit public discourse as well as individuals.

Forcing changes in ownership and making foreign social media platforms American-owned changes the nature of communicative apps in ways that affect speech. Social media platforms’ design and content choices have an enormous impact on communication. For example, platforms enable and classify participation differently. TikTok’s algorithm recommends third-party content based on how long users engage with videos and react to them. Those who access this infrastructure have a different communicative experience from those who use other platforms.

A new owner of a communicative app can order communication and association in new ways. The most widely used social media and, increasingly, AI platforms follow an infrastructural model that Julie Cohen describes as “platformized” infrastructure. This model enables greater centralized control than is typical for public infrastructure such as a public library system. Platformization places greater control in platform owners’ hands, permitting their ideological orientations and their susceptibility to markets, law, and political power to shape platforms’ decisions about what content and communication they permit and privilege.

Owners’ influence can be relatively subtle on social media platforms and AI-driven communicative platforms because they engage in what I am going to call ‘algorithmic content mediation,’ to distinguish it from editing and curation, both of which imply more precision of control over the output. Newspapers know exactly what they will publish, and their owners have to tell (human) editors what they cannot publish. In contrast, online platforms can use filters and prioritize some kinds of content over others, but they rarely know exactly what will emerge given the speed and scale of speech online. For example, when Elon Musk started changing Twitter’s (now X’s) algorithms, a menswear writer called Derek Guy ended up being promoted all over the platform for reasons that remain unclear. This is why it is helpful to think of algorithmic content mediation as “deal[ing] in probabilities, error rates, costs, and benefits.”

V. Foreign Ownership and Communicative Infrastructure

By stressing the value of foreign communicative platforms, I am not implying that they should not be regulated. This essay confines itself to demonstrating that foreign apps contribute more value to the American public sphere than regulatory conversations acknowledge. Regulating foreign ownership of communicative infrastructure should involve assessing how far foreign political and cultural influences enrich American public discourse. Foreign apps, with their distinct influences, offer new ways for people to engage with each other. Infrastructural apps like major social media platforms also evolve continuously. This ought to be taken into account by the Supreme Court when it assesses how far regulating foreign apps burdens speech.

TikTok Inc. v. Garland offers an illustration of what is lost when judges do not account fully for the burden on speech. The Supreme Court acknowledged that the First Amendment rights of American users were implicated in the use of TikTok. However, it underestimated the extent and nature of the relationship between these rights and major foreign platforms like TikTok by failing to account for TikTok’s infrastructural role, and how this role would be affected by transferring TikTok to American owners. In the Supreme Court’s defense, TikTok’s arguments in the case did not highlight its infrastructural role or influence on public discourse and confined itself to discussing how its users were affected.

TikTok enables Americans’ First Amendment rights in three significant, under-considered ways: (1) content from TikTok circulates to other platforms; (2) TikTok, as a foreign platform, offers Americans unique ways to communicate and engage; and (3) TikTok’s practices trigger changes in competing platforms. Thus, access to Chinese (and other foreign) apps supports Americans’ First Amendment rights.

Lawmakers, judges, lawyers, and others described TikTok as affecting only the rights of its 170 million registered American users rather than accounting for the entire burden on speech. The enormity of this burden becomes clear when one considers that TikTok affects the speech of more Americans than just its users, who make up a third of all American adults and over half of Americans under 30. TikTok permits these people to engage with more than a billion people across the world. It triggers changes in the cultural industries and the social media sector, as other companies, media, and content creators react to TikTok’s innovations.

As I was writing this essay, President Trump approved a TikTok transfer of ownership agreement. The arrangement did not take the form that most commentators and lawyers—and arguably the Supreme Court—had envisioned it might: namely, that the algorithm would be transferred when the ownership changed. Instead, the TikTok agreement arranged for ByteDance to license the content recommendation algorithm to TikTok’s new owners. However, I am going to discuss the Supreme Court’s decision in line with the initial expectations that ownership of the algorithm would change, as that allows an examination of the broader question of how ownership affects digital communicative infrastructure.

New owners have a visible impact on the development of platformized apps for the reasons I highlighted in the previous section. This is evident from Meta’s goals for WhatsApp. Although WhatsApp’s co-founders chose to sell the app to Meta (then Facebook), news reports suggest that they were unhappy with the new owners’ design choices.  Similarly, Elon Musk’s influence on X, previously known as Twitter, offers an extremely visible example of the impact of ownership changes on communicative platforms, including social media. A change in ownership means new priorities, values, and goals inform an app’s evolution, which implies that changes in ownership transform communication apps.

Access to a diversity of communicative apps, including foreign apps, is critical to Americans’ right to explore new ways of engaging with each other. The major Chinese platforms are currently the only platforms that offer Americans an alternative to discourse ordered by what is increasingly called the American “tech oligarchy.” Although there are good reasons to regulate foreign apps (and American apps that threaten national security), there is no reason for the regulation to be so expansive that it blocks Americans’ access to them instead of regulating their harmful effects or tendencies narrowly. In this regard, the European Union’s approach to foreign platforms might have been the better approach for American policymakers to adopt. Robert Post has argued that the European Union should treat Google, which is foreign-owned in the EU, as “an essential component of the communicative infrastructure necessary to sustain the public sphere” because of its status as a significant online platform. This argument could also apply to foreign communicative platforms in the United States: They enable Americans’ First Amendment rights. However, the dominant social media platforms in the United States and most of the English-speaking world have historically been American. This has left American regulatory discourse unprepared for the constitutional questions that foreign platforms raise. Policy aside, the Supreme Court would have been better able to appreciate the scale of speech rights at stake if the problem in TikTok Inc. v. Garland had been presented as a right to access communicative infrastructure. Framed correctly, the TikTok case might have been the Lamont of modern media.

VI. Adapting Lamont for Web-based Media

Lamont v. Postmaster General, a 1965 Supreme Court case, addressed whether an American citizen’s First Amendment rights were violated when the U.S. Post Office detained his copy of the Chinese news magazine Peking Review in 1963. In Lamont, the Supreme Court ruled that controlling the free flow of mail to the American public violates the First Amendment, just as controlling the free flow of ideas does. The law that was struck down required recipients of “communist propaganda” to take certain steps to receive it. This was an impediment (as opposed to an obstacle) that the Supreme Court interpreted as a violation of the First Amendment. Although Lamont is considered a case about the rights of American listeners, it could be framed as addressing the regulation of American communicative infrastructure, specifically, the U.S. Postal Service.

Recognizing listeners’ rights was pathbreaking at the time and remains a relatively underdeveloped thread of First Amendment jurisprudence and scholarship. The ideal progression from Lamont would have been if First Amendment jurisprudence discussed Americans’ right to access foreign radio and television. In the period before Lamont and the world wars, foreign-owned radio was restricted in the United States. The concern at the time, although it included foreign propaganda, was primarily that the spectrum was limited and that foreign entities might find a way to run signal interference with American military communication. The scarcity doctrine generally allowed for regulating broadcasting infrastructure in ways that were directly linked to the technological constraints of the limited spectrum. Legal norms regulating foreign ownership of broadcasting originated in these technological constraints and not in what might have been a helpful discussion of how threats to national security should be measured against Americans’ right to access foreign infrastructure.

TikTok Inc. v. Garland illustrates why thinking in terms of communicative infrastructure matters. TikTok Inc. v. Garland was about Americans’ access to foreign communicative infrastructure, even though it was never framed in this way for the court. Access to infrastructure affects access to ideas, and offers a sharper framing of what the justices might have been alluding to when they acted to protect the free flow of mail in Lamont. Freedom of speech depends on the technological infrastructure that enables widespread democratic participation and supports the system of free expression. Social media audiences participate even through what looks like passive consumption. For example, watching a video about climate change is engagement because when people watch these videos, they tell the algorithm about their own likelihood, as well as their demographic group’s likelihood, of watching other videos like it. On TikTok, Lamont’s passive listener becomes a participant.

Modern communicative infrastructure is complicated because, unlike the postal service, it is privately owned. Nevertheless, protecting communicative infrastructure, including online platforms, is necessary to protect First Amendment values. The First Amendment cannot protect free speech values without the institutions, infrastructure, and industries that promote these values. First Amendment law protects some kinds of communicative infrastructure, like the press and broadcast media, but fails to account for forms of communicative infrastructure that technology and online platforms create, although social media platforms are the dominant online communication model.

VII. Restricting TikTok 

The events leading to the TikTok restriction may be familiar to many readers. President Trump initiated the TikTok ban during his first term, and President Biden’s administration sealed it in place with the Protecting Americans from Foreign Adversary Controlled Applications Act (PAFACA). PAFACA targets TikTok and potentially any other “foreign adversary controlled applications” by permitting the executive to designate them for restriction. Both administrations treated TikTok as presenting national security concerns because it was Chinese owned and controlled and therefore subject to the Chinese government, which they feared could influence data collection and algorithmic content targeting.

The D.C. District Court, before which the statute was challenged, deferred to the legislature’s and executive’s characterization of the problem, which can be summarized as follows: (1) China is a serious cyber espionage threat to the United States; (2) China has made significant efforts to gather data about Americans for intelligence purposes; (3) China has gathered data from Chinese companies as well as foreign companies in which it makes investments or from which it purchases data sets; (4) Chinese laws require Chinese companies to provide China with full access to their data for criminal and security investigations; and (5) Chinese companies are not independent in the way in which American companies are. Additionally, the assistant director of the FBI asserted that (6) China tries to pre-position companies in the US that it can co-opt later and (7) engages in influence campaigns to counter and suppress opinions it objects to about China. Additionally, the US government concluded, based on TikTok Global’s responsiveness to Chinese demands to censor content, that (8) TikTok would comply if the Chinese government asked it to manipulate content in the United States, especially since China maintains “a powerful Chinese Communist Party committee embedded in ByteDance,” (internal quotation marks omitted) and TikTok is reliant on and integrated with ByteDance. Up to the point of the district court hearings, content-related concerns featured prominently among the justifications for banning TikTok.

In contrast, the Supreme Court focused on TikTok’s data collection, eliding the concerns about content and ignoring PAFACA’s regulation of expressive activity. Focusing exclusively on its data collection goal, the court argued that Congress would have passed the law even if it were not concerned about manipulative content. After framing data collection as the law’s goal, the Supreme Court argued that PAFACA regulated conduct rather than speech and applied the intermediate scrutiny test, not the strict scrutiny test. The Court concluded that PAFACA, as applied to TikTok, satisfied the intermediate First Amendment scrutiny test because its objective—which the justices saw as preventing a foreign adversary, China, from leveraging its control of ByteDance to capture TikTok users’ personal data—qualified as an important government interest. The justices also decided that PAFACA did not burden substantially more speech than necessary.

VIII. Why TikTok Inc. v. Garland is Troubling

The Supreme Court’s decision should concern everyone interested in speakers’ rights, listeners’ rights, healthy public discourse, and the rights of freedom of thought and association. The justices’ conception of social media platforms led them to underestimate the potential impact of divestment on First Amendment interests. The Supreme Court does not appear to have carefully considered how to protect expressive or associational activity in TikTok Inc. v. Garland: It treated TikTok as static, ignoring its role as an evolving, ubiquitous infrastructure. In addition to concluding that the statute did not burden substantially more speech than necessary, the Supreme Court argued that the statute’s “TikTok-specific provisions” requiring divestment of the company are facially content neutral, in the context of whether strict scrutiny applies. If the question before the justices had been framed as one of access to foreign infrastructure, different considerations might have been weighed against the national security interests animating PAFACA.

Major social media platforms, including TikTok, exhibit infrastructural features: they can be consumed non-rivalrously, enable downstream productive activity, and act as input into a wide range of goods and services. Their ownership plays a role in how they do these things. It is possible that the leasing of the TikTok algorithm from Douyin will ensure that TikTok’s unique infrastructural role is not lost. However, the Supreme Court did not anticipate or require this outcome. In other words, it did not consider how Americans’ First Amendment interests are served by access to foreign communicative infrastructure.

The Supreme Court’s assessment of whether speech was affected and to what degree was based on an outdated understanding of how web-based platforms and Chinese platforms affect Americans’ First Amendment interests. Communicative platforms often offer entirely new ways of communicating. Social media platforms shape how communication is “transmitted, received, and recirculated,” and the “experiences and perceptions of what is ordinary and natural in the linked realms of private and public interconnection and sensemaking.” They also shape which people engage with each other and how people interact with bots. By affecting online association, they influence communication.

The infrastructural properties of each platform are distinct. Some platforms privilege content that elicits reactions from others, and other platforms (e.g., Bluesky) avoid rewarding reactions, choosing to treat the sharing of someone else’s post in exactly the same way as annotated sharing of the same post. Platforms also construct associations differently (although their choices may affect those of their competitors); TikTok’s method of profiling users has influenced Meta. Safeguarding infrastructural diversity does not imply that platforms cannot be restricted or regulated, but shows that their differences play an important First Amendment role that should be considered while regulating them.

IX. Conclusion

The value of Americans’ right to access communicative infrastructure, including foreign communicative infrastructure, is under-considered in legal debate about communicative apps. Newcomers to the social media and artificial intelligence industries struggle to break through and are almost always purchased by the dominant firms, as WhatsApp and Instagram were. This leaves online communicative infrastructure very vulnerable to capture by private power. Focusing on citizens’ access to multiple sources of information and modes of engagement means that foreign communicative apps will remain an important source of alternative information, engagement, and competition. The judiciary’s willingness to defer to the legislative and executive branches’ characterization of communicative infrastructures as national security threats without careful consideration of their democratic value substantially undermines the First Amendment because of the scale of the speech and association affected.

If foreign-owned apps are easily restricted, communicative infrastructure will be left largely in the hands of a few American companies. Although these companies are not controlled by foreign states, they do have an interest in foreign markets and competing for profits. There is no reason to assume that they will never present or enable a national security threat. Using foreign ownership or control as a proxy for determining such a threat avoids the question of how communicative infrastructure should be regulated to mitigate national security threats and promote First Amendment values. Scholars have already proposed strategies, including fiduciary responsibility, interoperability, and anti-trust regulation, that might mitigate concerns about social media infrastructure. Layering additional regulations to address national security concerns is a more effective way to address concerns about foreign states than barring foreign-owned or foreign-controlled social media. Until American structural regulation improves, it is likely that apps capable of competing with dominant US communicative apps will only grow significantly in foreign countries with protected markets

Because non-US apps are shaped by foreign political, cultural, and economic forces, they create an environment for online engagement that American companies cannot create. They offer Americans what American companies have offered other publics in other countries for years: new ways to engage with ideas, with each other, and with people across borders. As communicative infrastructure, these apps are significant. To compel them to divest is to remove the quality that makes them unique: that their design reflects a distinct foreign influence.

Acknowledgments

I thank Maria Angel Arango, Jack Balkin, Katherine G. Bass, Liang Chen, Julie Cohen, Yuval Erez, Akriti Gaur, Jennifer Heinrichson, Margaret Jameel Jaffer, Hu, Shira Minsk, Michael McGovern, John Langford, Anat Leshnick, Lyrissa Lydsky, Yuping Lin, Dr. Margaret A. Oppenheimer and Cambridge Proofreading, John Peters, Robert Post, Alan Rozenshtein, Nishant Shah, David Schulz, Kalindi Vora, George Wang, Caleb Yong and Ethan Zuckerman. All mistakes are mine.

© 2026, Chinmayi Arun.

Cite as: Chinmayi Arun, The Right to Access Foreign Communicative Infrastructure, 26-4 Knight First Amend. Inst. (May 1, 2026), https://knightcolumbia.org/content/the-right-to-access-foreign-communicative-infrastructure [https://perma.cc/C8HP-RD6L].