On 14 and 15 October 2026, the same Eleventh Circuit panel of Judges Newsom, Luck and Tjoflat vacated the preliminary injunctions against Florida's HB 3 and Georgia's SB 351. Both states can now enforce their laws while the litigation continues. HB 3 bars accounts for children under 14 and requires parental consent for 14- and 15-year-olds. SB 351 requires age verification and parental consent for under-16s, with fines of up to $2,500 per violation. The panel held HB 3 content-neutral and subject only to intermediate scrutiny.
The strongest case for the laws
The states' argument deserves a fair statement. Compulsive-use features such as infinite scroll, autoplay and push notifications are designed to hold attention, and parents have few tools to counter them. HB 3 does not ban speech on a topic. It targets platforms that combine algorithmic feeds with those features, and the panel found that this targeting is likely content-neutral. Legislatures that want to protect children from engagement-maximising design have a legitimate interest, and courts have long let states set age-based rules for minors.
The Supreme Court's 2025 decision in Free Speech Coalition v. Paxton strengthens that case. Decided on 27 June 2025, it held that Texas's age-verification law for content obscene to minors triggers and survives intermediate scrutiny, because it only incidentally burdens adults' protected speech. The reasoning was that adults have no First Amendment right to access such content without first showing proof of age.
Why Paxton does not carry this far
That reasoning rests on a specific premise: the content in question is unprotected as to minors. Social media is different. It is a mixed space of news, politics, religion, art and conversation, nearly all of it fully protected for minors and adults alike. A law that conditions access to that whole space on proving your age, or on a parent's consent, burdens protected speech directly. Calling the rule content-neutral because it is framed around platform features does not change who loses access to what.
The Eleventh Circuit did acknowledge that the laws implicate the First Amendment. EFF's analysis reports that the Georgia panel conceded SB 351 mostly applies to platforms hosting protected speech. It still held that the challengers had not shown that unconstitutional applications substantially outweigh the valid ones. EFF argues that this standard is close to impossible to meet.
That standard comes from the Supreme Court's Moody v. NetChoice decision of 1 July 2024. The Court vacated both the Eleventh and Fifth Circuit judgments there because neither court had conducted a proper facial analysis, and it reaffirmed that a plaintiff must show a substantial number of a law's applications are unconstitutional, judged against its plainly legitimate sweep. Moody told lower courts to map what a law covers before striking it down. It did not say that a law touching mostly protected speech should be enforced in the meantime.
The practical effect is a procedural trap. Facial challenges are hard to win, and as-applied challenges need a record that only develops once the law is operating. So the laws take effect first, and the evidence of harm comes afterward.
The evidentiary gap
The panel's treatment of Florida's coverage definition shows the problem. EFF criticises the ruling for accepting that HB 3's trigger, where 10% of a platform's users under 16 spend two or more hours a day on it, shows addictiveness. Time spent is a poor proxy for harm. A teenager reading news, following a church group or practising a language for two hours is not showing a symptom. A rule built on usage thresholds will sweep in services that are plainly beneficial and may miss the ones lawmakers actually worry about.
SB 351's "commercially reasonable efforts" to verify age create a second problem. Verification applies to every user, so adults must prove their age too. That means handing identity documents or biometric estimates to platforms or third-party vendors, creating new data stores that become targets. A policy meant to protect children therefore adds privacy and security risk for everyone, and the record on whether it improves child safety is thin.
What proportionate regulation looks like
A pro-innovation approach would aim at the harm directly. Options include limits on design features for minors, such as default-off autoplay and notification curfews. Others are stronger default privacy settings, and device-level parental controls that families opt into without a universal ID check. These tools address the compulsive-use concern and leave teenagers' access to speech intact. They also avoid forcing every user through an age gate.
The litigation is not over. NetChoice and the Computer & Communications Industry Association (CCIA) said they are evaluating all of their options, and the cases return to the district courts. For now, the practical advice is to document what happens. EFF is asking affected users to record lost accounts, failed age checks and consent obstacles. That evidence matters because courts have required a record before they will weigh the speech costs, and the people best placed to supply it are the teenagers and parents living under these rules.
The larger lesson is about sequencing. If courts allow laws to take effect before testing their effects on protected speech, then the burden of proof falls on those who lose access. A better order would test narrow, well-evidenced measures first and expand only on the strength of that evidence.