Executive Order No. 13925: A Stop Sign on the Cyber Freeway

Executive Order 13925, signed on May 28, 2020 and titled “Preventing Online Censorship,” arrived hours after a dispute between the President and Twitter over labels the platform had attached to two of his tweets. The order announced an ambitious project: to narrow the liability shield that Section 230 of the Communications Decency Act extends to online platforms, and to do so largely through the administrative agencies. Whether an executive order can meaningfully redirect a statute that Congress wrote, and that courts have spent more than two decades construing, is the question the document raises but does not answer. The metaphor of a stop sign planted in the middle of a freeway is apt: the gesture is visible, but the traffic does not necessarily obey it.

What the order actually directs

The operative machinery of EO 13925 is indirect. It does not amend Section 230, because an executive order cannot. Instead it instructs the Secretary of Commerce, through the National Telecommunications and Information Administration (NTIA), to file a petition for rulemaking with the Federal Communications Commission. That petition asks the FCC to propose regulations clarifying when a platform’s decision to restrict access to content is not “taken in good faith” within the meaning of Section 230(c)(2)(A), and to clarify the relationship between the immunity in subsection (c)(1) and the good-faith condition in subsection (c)(2).

The order layers on additional levers. It directs federal agencies to review their advertising spending on platforms accused of viewpoint discrimination, asks the Federal Trade Commission to consider whether moderation practices amount to deceptive acts under Section 5 of the FTC Act, and revives a “Tech Bias Reporting Tool” for collecting complaints. Each of these is a pressure point rather than a command. None of them, individually, rewrites the statute; collectively they describe a coordinated effort to reinterpret it from outside Congress.

The Section 230 the order misreads

Section 230 contains two distinct protections, and the order tends to blur them. Subsection (c)(1) provides that a provider of an interactive computer service shall not be “treated as the publisher or speaker” of information supplied by another. Subsection (c)(2) separately shields good-faith decisions to restrict access to material the provider considers objectionable. Courts have generally read (c)(1) as the broad workhorse of platform immunity, applying it to moderation choices without requiring a separate good-faith showing.

That reading matters here. The order’s theory is that platforms forfeit protection when they moderate in bad faith, and that the FCC should police the line. But if (c)(1) supplies immunity for moderation independently of (c)(2)’s good-faith qualifier, then attaching new conditions to “good faith” reaches only a subset of the cases the order seems to target. The drafting assumes a statute structured around a single good-faith gate; the case law has not treated it that way.

The First Amendment runs the other way

The order frames large platforms as something like public forums and treats their moderation as censorship. The constitutional difficulty is that the First Amendment constrains the government, not private companies, and a platform’s choice about what speech to carry is itself an exercise of editorial judgment that the First Amendment protects. Compelling a private publisher to host or to refrain from labeling particular speech raises a compelled-speech problem; penalizing a platform for its editorial choices raises a retaliation problem.

Challengers to the order pressed exactly this inversion. Suits brought in 2020 by advocacy organizations, with the Electronic Frontier Foundation among counsel, argued that the order retaliated against protected editorial decisions and chilled speech by threatening regulatory and financial consequences. The claim was not that the President lacks views on platform moderation, but that an order designed to punish disfavored editorial choices collides with the very constitutional protection it invokes.

Separation of powers and the limits of the pen

An executive order operates within authority the President already holds; it cannot create new statutory meaning. EO 13925’s design concedes this by routing its substance through petitions and agency discretion rather than direct command. The FCC remained free to decline the NTIA petition, to seek comment without acting, or to conclude it lacked authority to issue the rule the order envisioned. The FTC’s enforcement priorities are likewise its own. The order could set an agenda; it could not guarantee an outcome.

That structural weakness is also the order’s shield against immediate judicial reckoning. Because so much of it depends on later agency action, courts confronting early challenges had to weigh whether the injury was ripe and whether plaintiffs had standing to sue an instrument that mostly asked other actors to act. The litigation thus turned as much on timing and justiciability as on the merits of the First Amendment claims.

What came next, and what remains unsettled

The trajectory after issuance illustrates the gap between proclamation and effect. The NTIA filed its petition in the summer of 2020, and the FCC’s chairman later signaled an intention to move forward with a rulemaking, invoking a separate statement by Justice Thomas, accompanying a denial of certiorari, that urged courts to revisit the breadth of Section 230 immunity. No binding rule emerged from that process, and the order was revoked in 2021. The underlying questions did not vanish with it.

Those questions persist because they are statutory and constitutional, not merely political. Whether Section 230(c)(1) immunizes moderation independently of good faith, whether an agency may interpret the provision, and how the First Amendment treats platform editorial control are durable problems that continue to surface in litigation and legislative proposals. The debate over how the law should treat online platforms also intersects with broader concerns about surveillance and data, themes this publication has taken up in its commentary on geofence warrants and the boundaries of digital privacy. Readers tracking how these disputes move through the courts can follow them through the publication’s case tracker and ongoing commentary.

EO 13925 is best understood not as a change in the law but as an attempt to steer it from the roadside. The cyber freeway it sought to slow is governed by a statute Congress enacted and courts interpret, and a stop sign placed by executive order has no power to halt that traffic on its own. The order’s lasting significance lies less in what it did than in how clearly it exposed the contested seams of Section 230 — seams that the next phase of this debate, wherever it unfolds, will have to address directly. This publication offers commentary and analysis, not legal advice.