Digital Censorship and the New Gatekeepers: Who Controls Online Speech?

Two 2024 Supreme Court decisions frame the American debate, and neither says what partisans on either side often claim. In Murthy v. Missouri, decided on 26 June 2024 by a 6–3 vote in an opinion by Justice Barrett, the Court held that two states and five social media users lacked standing to seek an injunction against federal officials over content moderation. It did not rule on whether the government’s contacts with platforms were unlawful; it found the plaintiffs had not shown that their harms could be traced to the government and fixed by an injunction.

In Moody v. NetChoice, decided unanimously on 1 July 2024, the Court vacated and sent back rulings on Florida and Texas laws that restricted how large platforms moderate content. Justice Kagan’s opinion said a platform’s choices about which messages to include in a feed can be protected editorial control. Read together, the decisions leave two live questions: when government pressure on a platform crosses into coercion, and how far states can dictate a platform’s curation.

The European Union’s Digital Services Act, Regulation (EU) 2022/2065, lets the Commission fine very large platforms up to 6 percent of worldwide annual turnover. On 5 December 2025 the Commission issued its first non-compliance decision under the Act, fining X €120 million for three transparency failings: a blue-check verification system judged deceptive, an inadequate advertising repository and restricted data access for researchers.

That decision matters for this file because it was not a takedown order. The penalties concerned transparency and design, which is how regulation of speech platforms usually works in practice: through duties to assess risk, disclose and give users remedies. The United Kingdom’s Online Safety Act 2023 follows a similar model, placing duties on services to manage illegal content and children’s safety under the oversight of the regulator Ofcom.

Most arguments about censorship collapse two separate questions. The first is who is acting: a government, a platform applying its own terms, or an infrastructure provider such as a payment processor, app store or host. The second is what is being done: deleting, demoting, demonetising, labelling or cutting off access. A government ordering removal raises a constitutional question in the United States, while a platform enforcing its own rules largely does not.

That is why the gatekeeper stack described above matters. Reach can be reduced without a single deletion, so the evidence for any claim needs to show the mechanism and the actor, not only the outcome.

Before accepting a claim that a particular view was suppressed, ask five questions: which actor made the decision, what rule or legal basis they cited, whether the decision is documented in a notice, transparency report, court filing or internal record, whether the affected party had a way to appeal, and whether the claimed effect on reach has been independently measured. A claim that cannot answer the first two is a grievance, not yet an evidenced case.