What the Words Were Always Going to Compel

The setup

Texas passed a child-safety law, the SCOPE Act, that placed duties on the platforms where people post and read each other’s messages.[1] The provision at issue is a monitoring-and-filtering requirement: a covered service must build and run a strategy to keep known minors from seeing material the statute lists as harmful, including content that promotes self-harm, eating disorders, substance abuse, or the sexual exploitation of children.[1] Two lawsuits followed, and a federal trial court blocked the filtering requirement before it could take effect. The state attorney general appealed, and the case reached the Fifth Circuit, one of the federal appeals courts that sit between the trial courts and the Supreme Court.[1]

The federal sentence at the center of the appeal is Section 230 of the Communications Decency Act, which says that no provider of an interactive computer service may be “treated as the publisher or speaker” of information provided by someone else.[2] For nearly thirty years that sentence has worked as a shield against lawsuits over individual posts. The question here was whether it also blocks a state from imposing a forward-looking duty to filter in the first place.[1]

The question

When a court reviews an order that pauses a law while a case proceeds, called a preliminary injunction, it asks mainly whether the challenger is likely to win in the end, and it decides the underlying legal questions without deference to the trial court.[1] So the real question was one of preemption, the principle that a valid federal law overrides a conflicting state one. Does Section 230 preempt a state duty that is framed not as liability for a bad post, but as an affirmative command to construct a filtering system?[1]

The answer was not obvious, because a second doctrine cuts the other way. Two years earlier, in Moody v. NetChoice, the Supreme Court held that a platform’s choices about what to carry are its own protected speech.[3] If curation is the platform’s own speech, one might think a law regulating that curation does not treat the platform as the speaker of anyone else’s words, and so escapes Section 230 entirely.[1]

The reasoning, walked

The majority began by refusing to be governed by labels. Drawing on its own recent cases, it asked not how the statute is styled but what the duty actually requires, and specifically whether compliance would force a company to monitor, alter, or remove content that other people wrote.[1] That move is the load-bearing one: it makes the substance of the obligation, not its statutory dress, decide the case.

Applied to the SCOPE Act, the requirement necessarily forces a service to watch and suppress third-party content before minors can see it, which is the publishing function the shield protects.[1] The court fixed the boundary with a compact formula it borrowed from a related decision: “publishers do not filter audiences; they filter content.”[1] Because the law makes platforms filter content, it treats them as publishers.

Texas answered that its duty is only a prospective compliance obligation, untied to any particular post, and so different in kind from the old liability suits. The court held the framing “is not conclusive.”[1] If form controlled, it reasoned, states “could simply re-encode the tort law otherwise preempted by Section 230 in the form of statutory obligations, accomplishing exactly the same effect.”[1] That is the anti-circumvention step, and it is what lets the shield reach a regulatory mandate and not just a damages claim.

The same line explains what survives. A separate requirement that platforms verify a user’s age was upheld, because verifying age only sorts audiences and does not require editing anyone’s words, a distinction the Supreme Court had already blessed.[4] Age-gating lives; mandated content screening does not.[1] As for the collision with Moody, the court declined to choose between the two doctrines, holding that the First Amendment protects a platform’s editorial choices from government regulation while Section 230 protects those same choices from civil liability, so that “the argument that we must choose between them presents a false choice.”[3]

One judge parted from the majority on exactly this point. Writing separately, Judge Ho would have let the state law stand, reasoning that Section 230 is triggered only when a provider is treated as the publisher of another’s content, and that a law about how a platform curates reaches the platform’s own speech instead.[1] If curation is first-party speech for the First Amendment, he argued, it is first-party speech for Section 230 too, so a duty to curate to a standard falls outside the shield.[1] The disagreement is not about outcomes at the edges but about which of two Supreme Court ideas controls the middle.

What turns on it

The decision converts Section 230 from a defense raised in a lawsuit into a bar against a class of state regulation, and it lands amid a wave of state child-safety statutes built on a duty of care.[1] For anyone building trust-and-safety systems, the audiences-versus-content line is a usable map: a state may compel age verification, but within this circuit it may not compel automated screening of what minors are shown.[1] The separate opinion reads as an invitation to the Supreme Court to say whether curation-based duties escape the shield, which means the fault line the majority papered over is likely to be tested again.[5]

The setup

A family fell out over who was entitled to run a trust after its creator could no longer manage her own affairs.[1] The woman had set up a revocable living trust, a legal container for property that its maker can change or dissolve during life, and over the years she revised who would take charge if she became incapacitated.[1] When incapacity came, a court-appointed conservator asked the Probate Court to name the rightful trustee, and that court chose one of two relatives and ordered the trust funds handed over to him.[1] The relative who lost took the matter up to the Superior Court, the trial court that hears these appeals in Connecticut, where no transcript of the probate proceeding existed.[1]

The question

The Superior Court, without holding a hearing, ruled on its own initiative from the written filings alone and reversed the Probate Court.[1] A Latin phrase describes that: the judge acted sua sponte, meaning on the court’s own motion rather than in response to anything a party requested.[1] The question on the next appeal was whether a probate appeal in which no record was made below may be decided on the papers, or whether the law requires a full trial first.[3]

The timing is what made the answer hard. Eleven months earlier the Connecticut Supreme Court had modernized this corner of practice in Rutherford v. Slagle, holding for the first time that ordinary summary judgment, the pretrial procedure for deciding cases with no genuine dispute of fact, is available in probate appeals.[2] A judge reading that decision could reasonably think paper-based resolution was now fair game.[1]

The reasoning, walked

The court started from the statute, which draws a two-track line.[3] When a record was made in the Probate Court, the appeal is decided on that record and is not a fresh proceeding.[3] The negative implication carries the weight: when no record was made, the appeal is instead a trial de novo, a Latin term meaning the matter is tried anew from the beginning.[1]

What a trial de novo demands came next, taken almost verbatim from the recent precedent. The Superior Court “sits as a probate court” and decides the issue afresh, ignoring the result reached below, and “the parties are entitled to a full hearing” at which they may present evidence that could have been offered in the Probate Court whether or not it ever was.[2] The whole point of that hearing, the court explained, is to cure the informality of probate practice by letting the parties build a real evidentiary record.[2]

Set against that standard, the error was plain and decided the case. The Superior Court had rendered judgment on the merits solely from its review of the papers and “never held a hearing at which the parties were afforded the opportunity to present argument or evidence.”[1] That procedure was neither a summary-judgment motion, which no one had filed, nor the trial the statute required, but a third thing the law does not allow.[1] The judgment could not stand, and the case was sent back for the trial that should have happened.[1] No judge dissented.

What turns on it

This is the first appellate gloss on the newly modernized rule, and it polices the boundary the earlier case opened.[4] Summary judgment is now a tool in probate appeals, but a judge may not reach past it to decide the merits on the papers without a motion, notice, or hearing.[1] The architecture is strange to anyone outside estates practice: appealing a probate decree can mean a brand-new trial in a different court that disregards the decision below, sometimes with the appealing party still carrying the burden of proof.[2]

Assessment of the reasoning

Both opinions reason from the same instinct, that what a legal rule truly requires is a matter of function rather than form, and both are stronger for refusing an easy label.[1] The platform case is the more exposed of the two. Its anti-circumvention move is persuasive on its own terms, yet it leaves the tension with the Supreme Court’s curation cases unresolved, and the separate opinion shows how a court taking the same precedents seriously could reach the opposite result.[1] The probate case is narrower and, within its lane, close to airtight, because the statute’s two-track text does most of the work and the recent precedent supplies the rest.[2] Read together, the pair shows appellate reasoning at its most characteristic, deciding less by weighing outcomes than by fixing what the words on the page were always going to compel. This is analysis of how the courts reasoned, not legal advice.