Section 230 Preempts State-Mandated Social Media Content Filtering for Minors; Speculative Third-Party “Chill” Cannot Establish Pre-Enforcement Standing

1. Introduction

The Fifth Circuit reviewed two consolidated, pre-enforcement challenges to Texas House Bill 18 (“H.B. 18”), also titled “The Securing Children Online through Parental Empowerment Act,” regulating certain social-media-like “digital service providers” (“DSPs”) under Texas Business & Commerce Code Chapter 509.

Two district-court preliminary injunctions largely blocked the monitoring/filtering requirement and (in the SEAT case) also enjoined the targeted-ads, unlawful-ads, and age-verification requirements. The Fifth Circuit’s decision resolves: (i) whether the SEAT plaintiffs had standing to seek pre-enforcement relief despite not being regulated DSPs; and (ii) whether the monitoring/filtering requirement is preempted by Section 230 of the Communications Decency Act.

2. Summary of the Opinion

3. Analysis

3.1. Precedents Cited (and how they drove the outcome)

A. Standing and pre-enforcement “chill” doctrine

• Murthy v. Missouri, 603 U.S. 43 (2024): This was the opinion’s central standing compass. The Fifth Circuit borrowed Murthy’s insistence that, when the plaintiff is not the direct target of regulation, standing requires a non-speculative showing that third parties will respond to government action in a way that creates a substantial risk of imminent injury— e.g., “a substantial risk that, in the near future, at least one platform will restrict the speech of at least one plaintiff in response to [the challenged law].” The court also used Murthy to reject broad “right to listen” theories untethered to a concrete, specific speaker-listener connection.
• Clapper v. Amnesty Int'l USA, 568 U.S. 398 (2013) and Louisiana v. Haaland, 86 F.4th 663 (5th Cir. 2023): These cases supplied the “highly attenuated chain of possibilities” warning. The court framed the SEAT plaintiffs’ theory as too conjectural: the law regulates DSPs; DSPs might over-filter; plaintiffs might be among those affected.
• Dep't of Com. v. New York, 588 U.S. 752 (2019) and Food & Drug Admin. v. All. For Hippocratic Med., 602 U.S. 367 (2024): Cited for the requirement that unregulated plaintiffs show regulated third parties “will likely react in predictable ways” and that the regulation of “someone else” is likely to cause the plaintiffs’ concrete injury.
• Tex. State LULAC v. Elfant, 52 F.4th 248 (5th Cir. 2022) and Laird v. Tatum, 408 U.S. 1 (1972): Used to distinguish objective, legally grounded chill from mere subjective chill. This distinction did the heavy lifting in rejecting the SEAT plaintiffs’ asserted self-censorship and anticipated censorship.
• Book People, Inc. v. Wong, 91 F.4th 318 (5th Cir. 2024): The district court had relied on Book People, but the Fifth Circuit distinguished it: the law there “directly proscribed the plaintiffs’ own conduct” and imposed a concrete “injurious dilemma” (compelled ratings/flagging or loss of sales). That contrast highlighted what was missing for SEAT: direct regulation or reliable evidence that DSPs would likely censor specific SEAT-related speech.
• Bantam Books, Inc. v. Sullivan, 372 U.S. 58 (1963): Distinguished as involving an already-operational government pressure campaign producing tangible distribution effects, not speculative downstream effects of a statute on independent private actors.
• Virginia State Board of Pharmacy v. Virginia Citizens Consumer Council, Inc., 425 U.S. 748 (1976) and Virginia v. American Booksellers Ass'n, 484 U.S. 383 (1988): Invoked to address “right to listen” standing. The Fifth Circuit treated these as contexts with tighter, more concrete listener-to-speaker or plaintiff-to-regulation connections than SEAT’s generalized desire to access broad categories of speech.
• Pool v. City of Houston, 978 F.3d 307 (5th Cir. 2020): Cited for the rationale behind special First Amendment pre-enforcement standing, but also for the limits: the plaintiff must show serious intent to engage in proscribed conduct and serious enforcement intent—harder where the law regulates third parties rather than the plaintiff.
• Susan B. Anthony List v. Driehaus, 573 U.S. 149 (2014): Provided the “certainly impending” / “substantial risk” vocabulary for imminence in pre-enforcement cases.

B. Section 230 preemption

• Doe v. MySpace, Inc., 528 F.3d 413 (5th Cir. 2008): The opinion’s foundational Fifth Circuit authority for Section 230’s “broad immunity” for claims stemming from publication of third-party content, including “monitoring, screening, and deletion of content.” The majority treated H.B. 18’s monitoring/filtering obligations as precisely the type of state-imposed publisher duty Section 230 blocks.
• Free Speech Coalition, Inc. v. Paxton, 95 F.4th 263 (5th Cir. 2024), aff'd, 606 U.S. 461 (2025): Supplied the doctrinal distinction the court applied: Section 230 preempts liability “reliant on the harm done by third-party content,” but not “liability purely based on whether plaintiffs comply with the statute” when compliance is not about publication decisions. The court leveraged this to contrast the porn-site age-verification statute (not preempted) with H.B. 18’s filtering mandate (preempted), emphasizing: “publishers do not filter audiences; they filter content.”
• A.B. v. Salesforce, Inc., 123 F.4th 788 (5th Cir. 2024): Provided the opinion’s operational test: “look … to what the duty at issue actually requires”—if it “would necessarily require an internet company to monitor[, alter, or remove] third-party content,” Section 230 bars it. The court treated § 509.053 as meeting that test by design.
• Zeran v. Am. Online, Inc., 129 F.3d 327 (4th Cir. 1997): Quoted via Fifth Circuit precedents for the breadth and purpose of Section 230 and its concern with chilling effects from publisher liability. The opinion used Zeran to reinforce that Section 230’s protection is intentionally sweeping.
• Force v. Facebook, Inc., 934 F.3d 53 (2d Cir. 2019) and Ricci v. Teamsters Union Local 456, 781 F.3d 25 (2d Cir. 2015): Cited for Section 230’s purpose “to keep government interference in the medium to a minimum” and for broad cross-circuit interpretive alignment.
• Google, Inc. v. Hood, 922 F.3d 212 (5th Cir. 2016): Used to connect Section 230 to “First Amendment values” driving the CDA, underscoring overlap between constitutional editorial discretion and statutory preemption.
• Medtronic, Inc. v. Lohr, 518 U.S. 470 (1996), United States v. Locke, 529 U.S. 89 (2000), and Egelhoff v. Egelhoff ex rel. Breiner, 532 U.S. 141 (2001): Addressed (and discounted) reliance on a presumption against preemption in areas of traditional state authority (child protection), emphasizing that clear federal intent can overcome any presumption.

C. Preliminary-injunction framework and constitutional avoidance

• Atchafalaya Basinkeeper v. U.S. Army Corps of Eng'rs, 894 F.3d 692 (5th Cir. 2018) and Mock v. Garland, 75 F.4th 563 (5th Cir. 2023): Set the standard of review and the four-factor preliminary-injunction test.
• Elrod v. Burns, 427 U.S. 347 (1976): Cited for the proposition that even minimal loss of First Amendment freedoms constitutes irreparable injury.
• City of Philadelphia v. New Jersey, 430 U.S. 141 (1977) and Sojourner T v. Edwards, 974 F.2d 27 (5th Cir. 1992): Anchored the sequencing principle: resolve statutory preemption before constitutional questions when possible.
• A.A. ex. rel. Betenbaugh v. Needville Indep. Sch. Dist., 611 F.3d 248 (5th Cir. 2010) and United States v. Lipscomb, 299 F.3d 303 (5th Cir. 2002): Expressed the constitutional-avoidance norm applied to decline reaching First Amendment merits after finding preemption.

3.2. Legal Reasoning

A. The SEAT plaintiffs: standing collapses under third-party speculation

The SEAT plaintiffs’ theory was structurally indirect: H.B. 18 regulates DSPs; the plaintiffs are users, speakers, listeners, and advertisers. Their alleged injury depended on DSPs changing moderation or ad practices in ways that would censor the plaintiffs or deprive them of content. The Fifth Circuit required (at the preliminary-injunction stage) a clear showing that standing is likely to obtain.

The court treated this case as presenting an “additional hurdle” beyond ordinary pre-enforcement standing because the asserted injury would be caused by “the independent action of some third party not before the court,” unless the plaintiffs could show DSPs would “likely react in predictable ways” and that such reactions would create a “substantial risk” of imminent suppression “at least partly in response to” H.B. 18. Applying that framework, the court found the record showed only conjecture: “may need” to over-filter, “could” be blocked, generalized fears about topics rather than specific, likely-to-be-suppressed speech, and listening injuries not tethered to concrete speaker-listener relationships.

Importantly, the court did not rule that unregulated parties can never have standing in this kind of case; it ruled that this record did not make the required clear showing. That is why the remedy was vacatur of preliminary relief and remand—not dismissal.

B. The CCIA plaintiffs: Section 230 preemption turns on what the duty requires

On the merits for CCIA/NetChoice, the panel began with preemption as a statutory ground that can (and should) resolve the case without reaching constitutional questions. The key move was characterizing § 509.053 not as a neutral compliance obligation unrelated to publication decisions, but as a duty that “would necessarily require an internet company to monitor[, alter, or remove] third-party content.”

The monitoring/filtering requirement compels DSPs to develop and implement strategies to prevent minors’ exposure to “harmful material” and other defined categories of content, and it supplies enforcement mechanisms by the Attorney General and by affected parents. In the majority’s view, this is paradigmatic publisher-role regulation: the State is setting “publication standards,” prescribing or suggesting moderation methods, and authorizing actions for failures to filter. Even if framed as an affirmative duty (rather than liability for a specific harmful post), the duty still targets the same editorial functions—monitoring and removal—and therefore triggers Section 230(c)(1)’s command that providers not be “treated as the publisher or speaker of any information provided by another information content provider.”

C. The court’s reconciliation of Section 230 with First Amendment “editorial discretion”

A notable doctrinal tension runs through the case: the Supreme Court in Moody v. NetChoice, LLC recognized that platforms exercise First Amendment-protected editorial discretion when curating feeds. Texas argued that if curation is a platform’s “own speech,” then Section 230 should not apply to state rules that effectively demand particular curation outcomes. The majority rejected the premise that the First Amendment and Section 230 are mutually exclusive: the same editorial acts can be constitutionally protected and statutorily immunized from being converted into liability for publishing others’ content.

This refusal to collapse Section 230 into the First Amendment is consequential: it preserves Section 230 as a distinct preemption/immunity rule that blocks state efforts to control content moderation through civil enforcement, even if the regulated moderation decisions might also be “speech” under the First Amendment.

D. Judge Ho’s partial dissent: “curation,” not “publication”

Judge Ho would have vacated the injunction in full, rejecting the majority’s preemption holding. His core claim: Section 230 is “triggered only” when a provider is treated as the publisher or speaker of content provided by “another,” and does not apply when the provider is held responsible for its own content—specifically, “its curation of third-party content.” He invoked: Doe ex rel. Roe v. Snap, Inc. (Thomas, J., dissenting), Anderson v. TikTok, Inc., A.B. v. Salesforce, Inc., Moody v. NetChoice, LLC, and Little v. Llano County.

The majority acknowledged the “specter” of overbroad immunity and discussed Anderson v. TikTok, Inc., but treated it as premised on a “false choice”: first-party speech protection and Section 230 preemption can “coexist.” The split reflects an emerging fault line in post-Moody Section 230 theory: whether algorithmic or moderation “curation” should be treated as the platform’s own “speech” for Section 230 purposes, narrowing preemption, or whether Section 230 remains focused on preventing state-law duties that force monitoring/removal of third-party content regardless of First Amendment framing.

3.3. Impact

A. Practical consequences for H.B. 18-like child-safety statutes

• Content-filtering mandates face a steep Section 230 obstacle in the Fifth Circuit when they functionally require monitoring/removal of user-generated content. Even careful drafting as “reasonable steps” or “strategies” may be treated as publisher-role duties.
• Age-verification regimes survive more easily (at least when patterned on the regime upheld in Free Speech Coalition v. Paxton, 606 U.S. 461 (2025)), because they regulate access mechanics rather than compelling editorial suppression of third-party content.

B. Litigation posture: who can sue, and when

• Platforms and regulated entities remain the most reliable plaintiffs for pre-enforcement challenges. Unregulated users, advertisers, and speakers may struggle unless they can produce concrete, near-term evidence that platforms will likely restrict their specific speech (or that they have a sufficiently concrete listener connection to identifiable speakers whose speech will be restricted).
• Record-building becomes decisive: the opinion’s standing rejection was heavily factual—what was “likely” versus what “may” happen. Future plaintiffs may attempt to cure standing on remand through clearer evidence of platform policy changes traceable to the statute, enforcement signals, or actual moderation incidents.

C. Doctrinal ripple effects: a live inter-circuit and intra-circuit debate

• Section 230 after Moody remains contested. The concurrence/dissent spotlights a growing interpretive dispute about “curation”: whether it narrows Section 230 (as Judge Ho suggests, consistent with Anderson v. TikTok, Inc.) or whether Section 230 continues to bar state laws that effectively compel content moderation of third-party speech (as the majority held).
• Potential Supreme Court interest increases where Section 230, child-safety regulation, and platform editorial discretion intersect—particularly given Justice Thomas’s concerns in Doe ex rel. Roe v. Snap, Inc. and the practical stakes for state regulation.

4. Complex Concepts Simplified

• Pre-enforcement challenge: A lawsuit filed before a law is enforced against the plaintiff, seeking to prevent enforcement based on predicted harm.
• Standing (injury, traceability, redressability): A plaintiff must show a real or imminent injury caused by the defendant that a court can fix. Here, SEAT’s difficulty was showing an imminent injury when the law regulates DSPs, not SEAT, and the injury depends on what DSPs might do.
• “Chilling effect” on speech: People self-censor to avoid feared consequences. Courts require an objective basis for the fear; a purely subjective chill is insufficient.
• Section 230(c)(1): A federal rule that prevents states from treating online services as the “publisher or speaker” of information provided by someone else. In this case, the majority viewed Texas’s filtering duty as forcing publisher-like moderation decisions about third-party posts.
• Preemption: When federal law overrides conflicting state law. If Section 230 preempts a state requirement, courts can enjoin it without deciding the First Amendment questions.
• “Curation” vs. “publication”: Judge Ho framed curation (algorithmic ordering, selection, and presentation) as the platform’s own speech; the majority treated the challenged duty as necessarily requiring monitoring/removal of third-party content—publisher functions Section 230 protects.

5. Conclusion

Computer & Comm v. Paxton establishes two practically significant rules for online-speech regulation litigation in the Fifth Circuit. First, unregulated plaintiffs alleging a First Amendment chill from a law aimed at platforms must do more than hypothesize overbroad platform reactions; they must make a clear, non-speculative showing of a substantial, imminent risk that platforms will restrict their speech (or a concretely connected speaker’s speech) because of the challenged law. Second, state statutes that mandate platform monitoring and filtering of user-generated content—however styled as reasonable, affirmative duties— are preempted by Section 230 when enforcement would necessarily require platforms to monitor, alter, or remove third-party content.

The court’s statutory resolution leaves the constitutional frontier largely untouched, but the concurrence/dissent underscores that Section 230’s scope—especially after Moody v. NetChoice, LLC—is in active doctrinal flux. The immediate effect is to block Texas’s principal content-filtering mechanism for minors under H.B. 18 (as challenged here), while simultaneously tightening the evidentiary and doctrinal path for users and advertisers to sue platforms’ regulators before enforcement.