from the the-censorship-industrial-complex? dept

Last year, the Marco Rubio-run State Department announced that it would start denying visas to people who worked in misinformation/disinformation research, content moderation, fact-checking, or other compliance and trust & safety roles. So, yeah, if you were an EU-based person who worked on preventing child sexual abuse material from appearing online, the US government decided you were not allowed in the country, bizarrely (and falsely) claiming you engaged in “censorship of protected expression.”

Except, as we’ve explained over and over again, that makes no sense. Only government officials or those working directly at the behest of the government can engage in censorship of protected expression. Otherwise it’s just private parties using their own rights of association to figure out what content they wish to associate with. And the actual reality (which MAGA culture warriors refuse to recognize) is that nearly all trust & safety work has fuck all to do with removing content. Much of it is literally about making platforms better and more trusted overall.

But, because the MAGA crew has been whipped up into a misinformation frenzy over the last decade that any research regarding mis- or disinformation is “censorship,” Trump and Rubio decided to throw the base some red meat and claim they were going to deny visas to people who worked in the field.

Thankfully, a federal court has pointed out that the only one engaging in censorship here is the Rubio State Department. By designating a group of people to be denied visas based on their own speech and association regarding disinformation research, the State Department engaged in unconstitutional viewpoint discrimination.

Judge Boasberg (who traditionally has been pretty conservative, but with Trump nonsense appearing before him quite frequently seems to now recognize that this administration is full of shit) notes how silly all of this is. After highlighting that most users of websites actually do want those sites to block scams and spam (what most content moderation is), though admitting that some people think of it as censorship, he points out that Rubio’s policy punishes researchers (the plaintiffs in this case) for trying to research and fact check disinformation.

The First Amendment reflects “a profound national commitment to the principle that debate on public issues should be uninhibited, robust, and wide-open.” N.Y. Times Co. v. Sullivan, 376 U.S. 254, 270 (1964). That commitment is not confined to stump speeches, editorials, or familiar forms of political advocacy….

Those principles cover the activity chilled here. CITR’s work depends on researchers who study how platforms structure public debate, report on misinformation and disinformation, advocate for access to platform data, petition officials, speak to the press, and collaborate with one another to set standards and press for reform. Some of that work culminates in reports, interviews, comments, petitions, and testimony. Some of it occurs before publication, in the candid exchange among researchers and organizations that makes public-facing work possible. Those activities, at least as reflected in this record, fall within the Amendment’s protection for speech, publication, petitioning, and expressive association. They also sit directly within the contested public debate over how online platforms structure discourse and whether, when, and how they should moderate harmful or false content….

CITR’s asserted injury is therefore not merely derivative of what its noncitizen members might say or what CITR might hear. The policy allegedly impairs CITR’s own work: who will contribute to its reports, what those reports can say, who will attach their names to them, and whether researchers will participate in the convenings and candid exchanges from which CITR’s public work emerges. See supra Section III.A.2.a. That is itself a First Amendment burden, as the Amendment protects both an organization’s creation and dissemination of information, Sorrell v. IMS Health Inc., 564 U.S. 552, 570 (2011), and the associational activity that makes collective speech possible….

The judge points out that the visa policy is already having an impact on this kind of research:

The reaction here was not merely predictable; it was all but ordered. Announcing the enforcement actions against leaders of two CITR member organizations, Rubio warned others engaged in the same work to “reverse course” or face the same. … The record shows that the message landed. Member A has refrained from international travel, including to CITR’s 2025 summit in Berlin, because of fear of being denied reentry under the policy; has limited public advocacy with CITR to a behind-the-scenes role because of fear of detention and deportation; and says that he or she would be substantially more likely to resume public association with CITR were the policy no longer in place…. Dr. Emma L. Briant, a U.K. citizen and Visiting Associate Professor at Notre Dame, likewise avers that the policy has caused her to self-censor in her writing and public speaking, hesitate to travel internationally, and evaluate even domestic speaking invitations against the risk of detention or deportation.

The judge calls out how the State Department started combing through visa applications to block “ordinary work” done by researchers and fact checkers, not limited to anyone actually engaged in any “censorship.”

The December cable supplies part of the answer. It directed consular officers to “thoroughly explore” visa applicants’ work histories, resumes, social-media profiles, and media appearances for involvement in “combatting misinformation, disinformation or false narratives, fact-checking, content moderation, compliance, and trust and safety,” and, on locating it, to pursue a finding of ineligibility…. Those categories do not describe the exercise of foreign sovereign power. They describe the ordinary work of researchers, fact checkers, platform employees, compliance officers, and nonprofit advocates who study, criticize, participate in, or press for content moderation. A cable that treats that work as evidence of immigration ineligibility reaches far beyond the coercive acts described in the May Memo: threats of arrest, payment freezes, legal compulsion, detention, fines, and demands for private data directed at American platforms or persons in the United States.

While the Court declines to review specific visa denials, it notes that the State Department clearly seems to be denying visas to people by claiming “censorship” when they had nothing to do with censorship. Indeed, the denials usually were about the State Department punishing people for First Amendment protected speech that the US government didn’t like. None of the justifications appear to actually be censorial:

The actions matter because State held them out as examples of the policy at work. Its public rationales identify the activity it treats as “complicity” in “censorship”: a report on hate speech and disinformation, advocacy directed at advertisers and platforms, disinformation-risk ratings, a petition for researcher access to platform data, a broadcast interview, and nonprofit leadership in organizations that help targets of online abuse seek removal of content aimed at them… Some of those justifications are tied to familiar First Amendment activity: reporting, speaking, petitioning, advocating for platform regulation, and associating through nonprofit leadership. Id. At least as to the private researchers and nonprofit leaders in CITR’s field, the public explanations do not identify any exercise of foreign sovereign power akin to the coercive acts the May Memo enumerates.

As Judge Boasberg notes, if you call all of that “complicity in censorship” then the term “censorship” has no real meaning:

If disinformation-risk ratings, reports on hate speech, petitions for platform-data access, advocacy, or nonprofit work seeking to limit abusive content can count as “complicity” in “censorship,” the policy has no clear stopping point short of the field itself — a concern sharpened by the Department’s announcement that it “stands ready and willing to expand” the list…. A lawful permanent resident working on a platform’s trust-and-safety team, a noncitizen researcher urging stronger disinformation labels, a compliance employee helping apply moderation rules, or an advocacy leader pressing advertisers away from sites that spread falsehoods could reasonably understand the policy to place their immigration status at risk — not because they wield foreign sovereign power or facilitate its censorship, but simply because they work in content moderation.

In its response, the DOJ pulled the usual MAGA nonsense of stomping its feet and just repeating “but content moderation is censorship” and making vague assertions about how these researchers aid foreign governments in censorship. The judge is not impressed.

The trouble is that the enforcement record does not honor that line. The Government has tied none of the private researchers and nonprofit leaders targeted in December to any exercise of foreign sovereign power. Pressed on that gap, counsel did not supply the missing connection. The Government stepped back from the five examples, explaining that it lacked “the full factual records or the reasons for those determinations” and that it would not be “fair to rely on those five” in gauging the policy’s scope…. But Defendants cannot publicly announce examples of the policy at work, warn that the Department stands ready to expand them, and then — when those examples prove inconvenient — deny that they reveal anything about the policy’s reach. A limiting principle that the Government cannot reconcile with its own enforcement record is no limit at all.

And thus, all this is classic, unconstitutional, viewpoint discrimination:

The policy, at its core, does not burden all speech about platforms, all research into content moderation, or all advocacy about online harms. It presses its enforcement thumb against one side of the scale: the view that platforms should do more to moderate content, label disinformation, restrict abuse, share data with researchers, or take responsibility for the harms their systems amplify. The Government, in other words, has not set itself against everyone who speaks about platform governance. It has set itself against those whose work favors more moderation rather than less. A noncitizen calling for less moderation, after all, has no comparable reason for concern under the policy.

Such action lies at the core of viewpoint discrimination. “At its most basic, the test for viewpoint discrimination is whether — within the relevant subject category — the government has singled out a subset of messages for disfavor based on the views expressed.” ….

The First Amendment does not permit officials to resolve that dispute by attaching legal burdens to the side they condemn

But that’s exactly what Rubio did here. If you worked on calling out disinformation, you could get your visa denied (or if you already had it, pulled). Judge Boasberg notes that if the policy were actually limited to foreign officials engaged in censorship, then the State Department might have an argument. But it’s not.

Even better, the court states that you can’t just call disinformation research “censorship” and pretend that’s a fact when reality says otherwise:

Protecting Americans from foreign officials who use sovereign power to suppress protected expression in the United States is in the Government’s interest. But the record does not show that the policy serves only that end. It instead brands a range of private expressive and platform-governance activity as “censorship,” without identifying any foreign-sovereign power that those actors exercised or helped exercise. The Government cannot make protected private expression a facially legitimate and bona fide basis for immigration consequences simply by placing it under the capacious and contested label of “censorship.”

He even notes that policy would violate the First Amendment under lower levels of scrutiny, meaning that even if the government could convince the court there was some justification for the policy, it still wouldn’t survive First Amendment scrutiny.

The judge doesn’t kill the policy entirely, noting that there may be cases where the State Department has a legitimate reason to deny a visa to someone actually engaged in efforts to silence Americans. It also hurts that when the Court sought evidence of visas being denied to actual censorship by sovereign officials, the State Department apparently came up empty:

The Department reportedly examined whether European regulators were using the Digital Services Act to censor American speech and found “no evidence that Member States of the European Union are overreaching the [Digital Services Act] to censor and criminalize online content.”

Oops! Sure, that goes against the narrative Rubio and MAGA folks have spun up about the EU being nothing but a bunch of censors, but when they can’t show the court any proof that they’re using this policy to go after actual government censors (while the plaintiffs can show where the policy was used to suppress or punish the speech of non-government censors) the end results are unlikely to make Rubio happy.

Measured against one another, the policy’s legitimate applications ultimately do not carry the day. The mismatch between Defendants’ asserted interest and the policy’s demonstrated operation is stark…. The defect identified above is not a feature of any one application; it is the policy’s selection criteria itself, and it travels wherever the policy does — into visa screening, exclusion, and removal alike. The policy’s legitimate applications, by contrast, remain episodic and largely undemonstrated. Whatever arithmetic might refine the comparison, the overbreadth inquiry asks whether a measure “prohibits a substantial amount of protected speech relative to its plainly legitimate sweep,” Hansen, 599 U.S. at 770, and a policy that selects its targets by an unconstitutional criteria, while its lawful uses remain occasional and largely unproven, answers that question.

Thus, the Court throws out this particular visa policy, though it doesn’t go quite as far as the researchers asked in requesting a protective order that would bar the government from using information related to this case in an immigration enforcement action. The judge recognizes that it’s still possible that the government could retaliate against these researchers, but hopes that this ruling will make them think twice about doing so. It also notes that if the government ramps up threats or actual retaliation against the researchers in this case, they can return to the courtroom to contest those actions.

For years, the loudest voices screaming about a “Biden censorship industrial complex” falsely insisted that pointing out disinformation was itself an attack on free speech. Now a federal court has found an actual, textbook case of unconstitutional censorship — carried out by the Secretary of State explicitly stripping visas from researchers based on their protected speech. I’m sure we’ll be hearing the same kind of outrage about Biden officials asking social media companies if they could be better in stopping health misinformation from spreading?

Filed Under: 1st amendment, censorship, chilling effects, content moderation, disinformation, free speech, james boasberg, marco rubio, research, state department, trust & safety, viewpoint discrimination, visas
Companies: citr