Opinion
1:24-CV-849-RP
08-30-2024
ORDER
ROBERT PITMAN, UNITED STATES DISTRICT JUDGE
Before the Court is Plaintiffs Computer & Communications Industry Association (“CCIA”) and NetChoice, LLC's (“NetChoice”) (collectively, “Plaintiffs”) Motion for a Preliminary Injunction. (Dkt. 6). Defendant Ken Paxton, in his official capacity as Attorney General of Texas, (“Paxton”) filed a response, (Dkt. 18), and Plaintiffs filed a reply, (Dkt. 24). Having considered the record, the parties' briefing, and the relevant law, the Court will grant in part and deny in part the motion for a preliminary injunction.
I. BACKGROUND
This case concerns Texas House Bill 18 (“HB 18” or “the Act”), a law that regulates social media websites. Tex. Bus. & Com. Code §§ 509.001, et seq. The Court will begin by describing the parties, then HB 18, and then the procedural posture of the case.
All cites are to HB 18, rather than to the Texas Business and Commerce Code.
A. The Parties
Plaintiff CCIA is a nonprofit organization that “promote[s] open markets, open systems, and open networks.” (Compl., Dkt. 1, at 4). Its members include major technology and social media companies such as Google (including YouTube), Meta (including Facebook, Instagram, and Threads), and X (formerly Twitter). (Id. at 4). Plaintiff NetChoice is a nonprofit trade association that comprises these members as well as the companies that own Nextdoor, Snapchat, and Pinterest. (Id.).
Paxton is the sole Defendant in this suit. He is the Attorney General of Texas, sued in his official capacity. (Id. at 5). Paxton and the Consumer Protection Division of the Attorney General's Office have enforcement authority under HB 18. See HB 18 § 509.151 (“A violation of this chapter is a deceptive act or practice actionable . . . solely as an enforcement action by the consumer protection division of the attorney general's office.”).
HB 18 also creates a separate private right of action. HB 18 § 509.152.
B. HB 18
1. Coverage
HB 18 regulates “digital service provider[s]” (“DSPs”) that “allow[] users to socially interact with other users.” HB 18 §§ 509.001(2), 509.002(a)(2). This definition includes standard social media platforms such as Facebook and Instagram as well as other “social” websites such as YouTube and Reddit. HB 18 defines DSPs as any “website,” “application,” “program,” or software that collects or processes personal identifying information (“PII”) and:
(1) “determines” both the “means” and the “purpose of collecting and processing the [PII] of users,” HB 18 § 509.001(1)-(2);
(2) “connects users in a manner that allows users to socially interact with other users on the digital service,” id. § 509.002(a)(1);
(3) “allows a user to create a public or semi-public profile for purposes of signing into and using the digital service,” id. § 509.002(a)(2); and
(4) “allows a user to create or post content that can be viewed by other users of the digital service, including sharing content on: (A) a message board; (B) a chat room; or (C) a landing page, video channel, or main feed that presents to a user content created and posted by other users,” id. § 509.002(a)(3).
Examples of covered DSPs include a “message board,” a “chat room,” and “a landing page, video channel, or main feed.” Id. § 509.002(a)(3). Exceptions to the law include: state and local government websites, financial institutions, medical websites, small businesses, institutions of higher education, employee management software, school education software, and email services that only provide “social” functions that are “incidental” to their main service. Id. § 509.002(b). Notably, the law also exempts DSPs that “primarily” feature “news, sports, commerce, or content primarily generated by the [website]” but “allows chat, comment, or other interactive functionality that is incidental to the digital service.” Id.
2. Requirements
a. Age Registration and Parental Disputes
With the stated purpose of protecting minors, HB 18 imposes age registration and verification requirements on users. Under HB 18, covered DSPs must make users “register” their age before those users can create an account. Id. § 509.051(a) (covered DSPs “may not enter into an agreement with a person to create an account with a digital service unless the person has registered the person's age with the website.”). A user cannot change their registered age “unless the alteration process involves a commercially reasonable review process.” Id. § 509.051(c).
HB 18 also allows parents to dispute the registered age of their child by notifying “a [DSP] that the minor is younger than 18 years of age” or successfully disput[ing] the registered age of the minor ....” Id. § 509.051(d)(2). If a parent disputes the age of their child, the DSP must treat the user as a known minor. Id. § 509.051(e). To ensure parenthood, HB 18 requires DSPs to “verify, using a commercially reasonable method and for each person seeking to perform an action on a digital service as a minor's parent or guardian: (1) the person's identity; and (2) the relationship of the person to the known minor.” Id. § 509.101(a).
The statute also allows parents to dispute the age of a minor if the parent “performs another function of a parent of guardian under this chapter.” HB 18 § 509.051(d)(2)(C). It is not entirely clear what this language refers to.
b. “Monitoring-and-Filtering” Requirements
If a known minor logs into a covered website, HB 18 imposes several new requirements on the DSPs. Chief among these are the requirements set forth in Section 509.053, which the Court refers to as the “monitoring-and-filtering” requirements. As this label suggests, these requirements make DSPs monitor certain categories of content and filter them from being on display for known minors. Specifically, HB 18 requires covered DSPs to “develop and implement a strategy to prevent [a] known minor's exposure to” defined categories of prohibited speech. Id. § 509.053(a). HB 18 prohibits content “that promotes, glorifies, or facilitates” the following categories:
(A) “suicide, self-harm, or eating disorders”;
(B) “substance abuse”;
(C) “stalking, bullying, or harassment”;
(D) “grooming, trafficking, child pornography, or other sexual exploitation or abuse”; and
(E) material that qualifies as obscenity for minors under Texas Penal Code § 43.24.Id.
Strategies to prevent minors' exposure to these harmful categories “must include”:
(A) creating and maintaining a comprehensive list of harmful material or other content described [above] to block from display to a known minor;
(B) using filtering technology and other protocols to enforce the blocking of material or content on the list [above];
(C) using hash-sharing technology and other protocols to identify recurring harmful material or other content described [above];
(D) creating and maintaining a database of keywords used for filter evasion, such as identifiable misspellings, hash-tags, or identifiable homoglyphs;
(E) performing standard human-performed monitoring reviews to ensure efficacy of filtering technology;
(F) making available to users a comprehensive description of the categories of harmful material or other content described [above] that will be filtered; and
(G) except as provided by Section 509.058, making available the digital service provider's algorithm code to independent security researchers.Id. § 509.053(b)(1).
In addition to the above mandatory strategies, HB 18 allows DSPs to use other strategies, including:
(A) creating engaging a third party to rigorously review the digital service provider's content filtering technology;
(B) participating in industry-specific partnerships to share best practices in preventing access to harmful material or other content described [above]; or
(C) conducting periodic independent audits to ensure:
i. continued compliance with the digital service provider's strategy; and
ii. efficacy of filtering technology and protocols used by the digital service provider.HB 18 § 509.053(b)(2).
c. Additional Requirements
In addition to the “monitoring-and-filtering” requirements, HB 18 imposes regulations on how DSPs can treat the data of known minors. First, the law imposes limits on what kind of data a DSP may collect. Covered DSPs must “limit collection of the known minor's [PII]to information reasonably necessary to provide the digital service” and “limit use of the [PII] to the purpose for which the information was collected ....” HB 18 § 509.052(1). In addition, a covered DSP may not: (A) allow the known minor to make purchases or engage in other financial transactions through the digital service; (B) share, disclose, or sell the known minor's PII; (C) use the digital service to collect the known minor's precise geolocation data; or (D) use the digital service to display targeted advertising to the known minor. Id. § 509.052(2).
Second, HB 18 requires DSPs to create and provide parental tools for digital services. Id. § 509.054. Parents must be able to “control the known minor's privacy and account settings,” alter data collection and financial transaction settings for the minor, and “monitor and limit the amount of time the . . . known minor spends” using the DSP. Id. § 509.054(b).
Third, HB 18 requires DSPs to “make a commercially reasonable effort to prevent advertisers on the digital service provider's [website] from targeting a known minor” with content that “facilitates, promotes, or offer[s] a product, service, or activity that is unlawful for a minor in this state to use or engage in.” Id. § 509.055.
Fourth, DSPs that publish content, more than one-third of which is obscene for minors under Texas law, “must use a commercially reasonable age verification method to verify that” the user seeking to access the information is 18 or older. Id. § 509.057. Plaintiffs do not challenge this provision.
Fifth, HB 18 requires covered DSPs that “use[] algorithms to automate the suggestion, promotion, or ranking of information to known minors” to “disclose . . . in a clear and accessible manner, an overview of”: (1) “the manner in which the digital service uses algorithms to provide information or content”; (2) “the manner in which algorithms promote, rank, or filter information or content”; and (3) “the [PII] used as inputs to provide information or content.” Id. § 509.056(2). DSPs that “use[] algorithms to automate the suggestion, promotion, or ranking of information” must “make a commercially reasonable effort to ensure that the algorithm does not interfere with the [DSP's] duties.” Id. § 509.056(1).
HB 18 contains various other provisions related to the practical enforcement of the law, such as clarifying that PII may be shared to comply with legal investigations, HB 18 § 509.059(1); that DSPs need not disclose trade secrets, id. § 509.058; and that the Department of Family and Protective Services may designate its staff as a minor's caregiver, if a minor is in the agency's conservatorship. Id. § 509.104.
C. Procedural History
Plaintiffs filed suit on July 30, 2024, and filed a motion for a preliminary injunction concurrently with their complaint. (Compl., Dkt. 1; Mot. Prelim. Inj., Dkt. 6). Paxton filed a response, (Dkt. 18), and Plaintiffs filed a reply, (Dkt. 24).
The parties filed a joint notice that the motion can be decided solely on the papers without a hearing. (Dkt. 19).
Plaintiffs allege that HB 18 violates the First Amendment because it is a contentbased law that does not survive strict scrutiny. (Compl., Dkt. 1, at 21-25). Their preliminary injunction argues both that the coverage provision and thereby the law as a whole is contentbased, and it argues that the monitoring-and-filtering requirements are independently content-based. (Mot. Prelim. Inj., Dkt. 6, at 14-24). Plaintiffs also argue that the law is void for vagueness, (Compl., Dkt. 1, at 25-26, 32-34), imposes unconstitutional prior restraints, (Id. at 26-32), and is preempted under Section 230 of the Communications Decency Act (“Section 230”). (Id. at 34-35). Plaintiffs seek a preliminary injunction of the law, requesting that the Court enjoin Paxton from enforcing HB 18 in full. (Mot. Prelim. Inj., Dkt. 6, at 20).
Paxton filed a response in opposition to the motion. (Dkt. 18). Paxton argues that Plaintiffs' claim is barred by sovereign immunity and that Plaintiffs lack standing to bring suit. (Id. at 18-23). Paxton also claims that HB 18 should not be classified as a whole and that its disparate provisions, many of which are not content based, should survive First Amendment scrutiny, to the extent the First Amendment applies at all. (Id. at 23-30). Paxton further argues that HB 18 does not violate the First Amendment, is subject to limiting constructions, is not unconstitutionally vague, and is not preempted under Section 230. (Id. at 30-44). Finally, Paxton notes four “issues” with Plaintiffs' motion: (1) the law is severable; (2) the law is not a prior restraint; (3) age verification is not challenged in Plaintiffs' motion; and (4) Plaintiffs improperly delayed bringing suit. (Id. at 44-45). Plaintiffs filed a reply, (Dkt. 24), and Paxton filed an opposed motion for leave to file a surreply. (Dkt. 23).
II. LEGAL STANDARD
A preliminary injunction is an extraordinary remedy, and the decision to grant such relief is to be treated as the exception rather than the rule. Valley v. Rapides Par. Sch. Bd., 118 F.3d 1047, 1050 (5th Cir. 1997). “A plaintiff seeking a preliminary injunction must establish that he is likely to succeed on the merits, that he is likely to suffer irreparable harm in the absence of preliminary relief, that the balance of equities tips in his favor, and that an injunction is in the public interest.” Winter v. Nat. Res. Def. Council, Inc., 555 U.S. 7, 20 (2008). The party seeking injunctive relief carries the burden of persuasion on all four requirements. PCI Transp. Inc. v. W. R.R. Co., 418 F.3d 535, 545 (5th Cir. 2005).
III. DISCUSSION
As Paxton notes, Plaintiffs' motion turns on the likelihood of success on the merits. (Resp., Dkt. 18, at 45 (“The remaining preliminary injunction factors will not meaningfully impact this case.”)). The Court will begin its likelihood of success analysis with the jurisdictional issues: whether Paxton is entitled to sovereign immunity and Plaintiffs' standing to bring suit. Finding that Ex parte Young applies and Plaintiffs have standing, the Court will then turn to the merits of Plaintiffs' motion.
A. Sovereign Immunity
The Eleventh Amendment deprives federal courts of jurisdiction over “suits against a state, a state agency, or a state official in his official capacity unless that state has waived its sovereign immunity or Congress has clearly abrogated it.” Moore v. La. Bd. of Elementary & Secondary Educ., 743 F.3d 959, 963 (5th Cir. 2014). However, lawsuits may proceed in federal court when a plaintiff requests prospective relief against state officials in their official capacities for ongoing federal violations under the Ex parte Young exception. 209 U.S. 123, 159-60 (1908). For the exception to apply, “the state official, ‘by virtue of his office,' must have ‘some connection with the enforcement of the [challenged] act, or else [the suit] is merely making him a party as a representative of the state, and thereby attempting to make the state a party.'” City of Austin v. Paxton, 943 F.3d 993, 997 (5th Cir. 2019) (quoting Young, 209 U.S. at 157).
The Fifth Circuit's test for a defendant with “some connection” to enforcement looks to three factors: (1) the particular duty to enforce the challenged law; (2) a demonstrated willingness of enforcement; and (3) enforcement authority that amounts to “compulsion” or “constraint.” Mi Familia Vota v. Ogg, 105 F.4th 313, 325 (5th Cir. 2024). Plaintiffs argue that all three elements are met because the Texas Attorney General's Office enforces HB 18, has shown a willingness to enforce the law, and the enforcement power constitutes compulsion and constraint. (See Mot. Prelim. Inj., Dkt. 6). Paxton focuses his briefing on the particular duty to enforce and demonstrated willingness to enforce. (See Resp., Dkt. 18).
As to his “particular duty,” Paxton suggests there is a distinction between a “particular duty to enforce the statute” and a mere “[d]iscretionary authority to act[.]” (Id. at 18 (citing Tex. All. for Retired Ams. v. Scott, 28 F.4th 669, 672 (5th Cir. 2022); Mi Familia Vota, 105 F.4th at 327)). Because HB 18 authorizes the Attorney General to enforce the law-but does not require enforcement against all known violations-Paxton argues that his authority does not exceed the discretionary authority to act.
Paxton conflates two different elements of the Ex parte Young test. Whether the defendant has the “particular duty” to enforce the challenged law is different than whether there is a “demonstrated willingness” of enforcement. See Mi Familia Vota, 105 F.4th at 325. To demonstrate a “particular duty,” Ex parte Young requires neither a history of enforcement nor a statutory requirement of enforcement. City of Austin, 943 F.3d at 1001; Air Evac EMS, Inc. v. Tex. Dep't of Ins., Div. of Workers' Comp., 851 F.3d 507, 519 (5th Cir. 2017); Tex. Democratic Party v. Abbott, 961 F.3d 389, 401 (5th Cir. 2020). Under HB 18, Paxton has the “particular duty” to enforce the law because a “violation [is] actionable . . . solely as an enforcement action by the consumer protection division of the attorney general's office.” HB 18 § 509.151. Whether that enforcement is mandatory or discretionary speaks to a “demonstrated willingness” to enforce the statute-not whether Paxton has the “particular duty” of its enforcement. See Natl. Press Photographers Assn. v. McCraw, 90 F.4th 770, 786 (5th Cir. 2024) (finding state agency heads with discretionary enforcement power had a particular duty to enforce and noting that “[a]s heads of Texas law-enforcement agencies, [defendants] have more than just the general duty to see that the state's laws are implemented-they are directly responsible for enforcing Texas's criminal laws”).
HB 18 also creates a private right of action for equitable relief, but Paxton is the sole governmental agent authorized to bring suit. HB 18 § 509.151-52.
As to the “demonstrated willingness,” there need be only a “scintilla of enforcement by the relevant state official” for Ex parte Young to apply. City of Austin, 943 F.3d at 1002 (quotations omitted). Imminent enforcement or actual threats thereof are “not required.” Air Evac, 851 F.3d at 519. Paxton suggests that he “must have taken some step to enforce HB 18” in order to be a proper Ex parte Young defendant. (Resp., Dkt. 18, at 18 (quoting Mi Familia Vota, 105 F.4th at 329)). But Plaintiffs bring a pre-enforcement challenge for a law that has not yet taken effect. Evidence of actual enforcement is therefore impossible and is not required. See Calhoun v. Collier, 78 F.4th 846, 851 (5th Cir. 2023). All Plaintiffs must show is “‘some scintilla' of affirmative action by the state official ....” Id. (citing Tex. Democratic Party, 961 F.3d at 401).
That “scintilla” is met. Critically, Paxton has not disavowed enforcement. See Virginia v. Am. Booksellers Ass'n, Inc., 484 U.S. 383 (1988) (“We are not troubled by the pre-enforcement nature of this suit. The State has not suggested that the newly enacted law will not be enforced, and we see no reason to assume otherwise.”); Babbitt v. United Farm Workers Nat. Union, 442 U.S. 289, 302 (1979) (“[T]he State has not disavowed any intention of invoking the criminal penalty ....Appellees are thus not without some reason in fearing prosecution for violation of the ban ....”). In the First Amendment context, this is enough. See Nat'l Press, 90 F.4th at 782 (“Unlike in other constitutional contexts, in the speech context, we may assume a substantial threat of future enforcement absent compelling contrary evidence.”) (internal citation and quotation omitted).
Beyond his refusal to disavow, Paxton's response to the preliminary injunction motion makes clear that he does intend to enforce the law. Paxton's motion repeatedly emphasizes that he believes social media is severely harming children. Paxton's introduction says, “[Plaintiffs] appear to think the Constitution leaves Texas powerless to defend its children . . . against the obvious dangers linked to social media use. They are wrong.” (Resp., Dkt. 18, at 11). Paxton compares social media sites to “a slot machine, . . . giv[ing] users dopamine hits every so often . . . to keep them hooked just a little longer.” (Id. at 13). Paxton's motion states that “[s]ocial media is hurting our children at an alarming and growing rate.” (Id. at 16). These statements are, at the very least, a “scintilla” of willingness to enforce. Any reasonable reader of these statements would consider them to indicate an intent to enforce. Paxton's refusal to disavow enforcement, combined with his strident criticism of the impacts of social media on children, shows a “scintilla” of willingness to enforce HB 18.
Paxton instead argues that a recent Fifth Circuit case-Mi Familia Vota-upended Supreme Court precedent on pre-enforcement challenges by requiring a mandatory (rather than discretionary) enforcement provision. (Resp., Dkt. 18, at 18 (citing Mi Familia Vota, 105 F.4th at 327)). In Mi Familia Vota, a group of plaintiffs challenged Texas's general election laws by suing, (among others), the Harris County District Attorney Kim Ogg. Id. at 317. Because Ogg had only the general duty to enforce the laws of the state, the Fifth Circuit found that the “[d]iscretionary authority to act, on its own, is insufficient to give rise to a particular duty to act,” which was “derived from the Texas constitution . . . and from the Texas Code of Criminal Procedure.” Id. at 327-28. The court also noted that Ogg “never enforced the challenged provisions in the past” and that Plaintiffs had not shown any “action taken by Ogg to show a demonstrated willingness to enforce.” Id. at 330. Indeed, Ogg stipulated that she would not enforce the law during the pendency of the suit. Id. Mi Familia Vota is readily distinguishable in at least three material ways. First, as just discussed, Paxton has shown a willingness to enforce. By describing at length the ways in which social media is harming children and how the law represents a substantial and compelling state interest, Paxton has shown a “scintilla” of willingness to enforce. Second, whereas Ogg stipulated that she would not enforce the law during the pendency of suit, Paxton has not done so here. Third, unlike Ogg's enforcement power in Mi Familia Vota, Paxton's authority to enforce HB 18 is not derived from the general constitutional duties of the Texas Attorney General. Instead, his enforcement authority is designed from the statute itself, vesting him with the “sole[]” power to take “enforcement action” under the law. HB 18 § 509.151.
Paxton' final argument suggests that Ex parte Young does not apply because the “consumer protection division of the attorney general's office” is vested with the authority to bring suits, rather than Paxton himself. (Resp., Dkt. 18, at 18). Ex parte Young requires only “some connection” to enforcement. Lewis v. Scott, 28 F.4th 659, 663 (5th Cir. 2022). “Where a state actor or agency is statutorily tasked with enforcing the challenged law and a different official is the named defendant, [the] Young analysis ends.” City of Austin, 943 F.3d at 998 (emphasis added). Paxton, as Attorney General, is in charge of the Consumer Protection Division of the Attorney General's Office. He “enforces the law” through its subdivisions, including the Consumer Protection Division. Book People, Inc. v. Wong, 91 F.4th 318, 335 (5th Cir. 2024). Paxton does not deny that he has ultimate control over the management of divisions in his office and the prosecution of cases brought by those divisions. That authority establishes “some connection” to the law. Because Paxton has the particular duty to enforce HB 18 and a demonstrated willingness to do so, he is a proper defendant under Ex parte Young. Accordingly, Plaintiffs' suit is not barred by sovereign immunity.
B. Standing
Next, the Court addresses Plaintiffs' standing to bring suit. Paxton raises several arguments on this point: (1) Plaintiffs lack pre-enforcement standing, especially on provisions that regulate activities Plaintiffs do not engage in; (2) Plaintiffs have no associational standing for their facial challenges; (3) Plaintiffs lack associational standing for their as-applied challenges; (4) Plaintiffs' as-applied challenge is not ripe; and (5) Plaintiffs lack standing to assert third-party rights of their users. (Resp., Dkt. 18, at 19-23).
1. Pre-Enforcement Standing
“In a pre-enforcement challenge, Plaintiffs can establish an injury in fact if they show that ‘(1) they have an intention to engage in a course of conduct arguably affected with a constitutional interest, (2) their intended future conduct is arguably proscribed by the policy in question, and (3) the threat of future enforcement of the challenged policies is substantial.'” Book People, 91 F.4th at 335 (quoting Speech First, 979 F.3d at 330). Paxton does not dispute these second and third prongs. (Resp., Dkt. 18, at 19). Instead, Paxton argues that Plaintiffs (and their members) lack standing to challenge many provisions of HB 18 because they do not engage in all activities regulated by the law, such as collecting a user's precise geolocation data and selling a minor's PII. (Id.).
Paxton correctly notes that “Plaintiffs must show an intent to engage in the conduct being regulated for each claim they press.” (Id. (emphasis in original) (citing TransUnion LLC v. Ramirez, 594 U.S. 413, 431 (2021); Susan B. Anthony List v. Driehaus, 573 U.S. 149, 159 (2014))). But Plaintiffs state in their complaint that “HB 18 mandates that covered websites adopt many costly requirements” and that the “haphazard” regulations likely “burden[] speech dissemination.” (Compl., Dkt. 1, at 3). Those regulations plausibly target at least some of Plaintiffs' members, and, at any rate, the Court does not enjoin HB 18's enforcement as to these other provisions. See infra, Section III.C.2.A.
2. Associational Standing
An association has standing to bring suit on behalf of its members when: “(a) its members would otherwise have standing to sue in their own right; (b) the interests it seeks to protect are germane to the organization's purpose; and (c) neither the claim asserted nor the relief requested requires the participation of individual members in the lawsuit.” Assn. of Am. Physicians & Surgeons, Inc. v. Texas Med. Bd., 627 F.3d 547, 550 (5th Cir. 2010) (citing Hunt v. Wash. St. Apple Adver. Comm'n, 432 U.S. 333, 343 (1977)).
Plaintiffs meet these three criteria. Beginning with the second prong, which is effectively uncontested, Plaintiffs seek to protect an interest germane to their organizational purposes. Both Plaintiffs are trade associations that aim to promote online commerce, speech, and accessibility for internet services. (Compl., Dkt. 1, at 4-5). HB 18 potentially interferes with that mission, regulating how minors can access their online platforms.
Paxton appears to focus his arguments on the first prong: whether Plaintiffs' members would have standing to sue in their own right. (See Resp., Dkt, 18, at 19). Here, Paxton relies heavily on the Supreme Court's recent decision in Moody v. NetChoice, LLC, 144 S.Ct. 2383 (2024). In Moody, the Supreme Court dealt with separate challenges to new Texas and Florida social media laws that regulated online platforms' ability to engage in content moderation. Id. at 2393. Rather than definitively rule on the constitutionality of the laws, the Supreme Court vacated the circuit court decisions, finding that the lower courts had failed to conduct proper facial analyses of the laws at issue. Id. at 2397, 2409. While Moody certainly cautions lower courts to treat facial challenges carefully, Moody is not a case about standing. The majority opinion does not contain a single mention of “standing,” much less third-party or associational standing. See id. Put simply, Moody did not alter the standing analysis for associations, only the proper scope of relief for when those associations challenge laws under the First Amendment.
Moody aside, there is little question that Plaintiffs' members would have standing to sue in their own right. Many of the members are social media companies and digital service providers who are regulated by HB 18; including Meta (Facebook and Instagram); Google (YouTube); Snap, Inc. (Snapchat); Nextdoor; Pinterest; and X. As objects of the regulation, those members would have standing to sue. Texas v. EEOC, 933 F.3d 433, 446 (5th Cir. 2019) (“If, in a suit ‘challenging the legality of government action,' ‘the plaintiff is himself an object of the action . . . there is ordinarily little question that the action or inaction has caused him injury, and that a judgment preventing or requiring the action will redress it.'”) (quoting Lujan v. Defs. of Wildlife, 504 U.S. 555, 561-62 (1992)).
Finally, turning to the third prong of associational standing, Plaintiffs' members do not need to participate in this suit. The inquiry regarding the participation of members does not derive from Article III but is “solely prudential.” Assn. of Am. Physicians & Surgeons, Inc., 627 F.3d at 550. The inquiry focuses importantly on “matters of administrative convenience and efficiency” and is assessed by “examining both the relief requested and the claims asserted.” Id. at 551. Paxton argues that Plaintiffs' “facial claims require the participation of individual members,” again suggesting that Moody impliedly changed this test. (Resp., Dkt. 18, at 12 (“Moody requires a member-specific analysis for facial claims in this context. Plaintiffs' members will need to participate in this lawsuit to resolve these facial claims, which means Plaintiffs lack standing to sue in their associational capacity.”)). To repeat, Moody's majority opinion does not mention standing or the requirement of individual participation. 144 S.Ct. at 2394. Moody does show that facial challenges may sometimes require individualized or detailed inquiries into members' activities, but that does not automatically necessitate those members' participation as parties in the suit. Other mechanisms such as detailed briefing and discovery will likely provide sufficient information to make individualized inquiries where needed.
If a trade association brings suit but cannot show in sufficient detail that a law is unconstitutional in the substantial majority of its applications, then it will not obtain a facial injunction. That is to say, the problem identified by Moody can be remedied by denying facial challenges on the merits, not dismissing the entire suit for lack of standing.
With the exact same Plaintiffs as in Moody, there is no prudential reason why Plaintiffs' members must participate in this suit. The preliminary injunction motion does not involve heavily contested issues of fact. This is especially true for the “monitoring-and-filtering” requirements, whose constitutionality does not seriously turn on Plaintiffs' members' particular activities. As is the case here, the parties' briefing and discovery will likely provide sufficient information to make individualized inquiries where needed. Finally, courts regularly allow membership organizations and trade associations to bring suit on behalf of their members when they seek to enjoin enforcement of a statute or regulation, rather than damages. See United Food and Com. Workers Union Loc. 751 v. Brown Grp., 517 U.S. 544, 553-54 (1996). Even acknowledging the “variegated” nature of the internet, individual participation is not required.
3. Association Standing for As-Applied Challenges
Paxton next suggests that Plaintiffs cannot bring an as-applied challenge on behalf of their members because an as-applied challenge would still require the participation of individual members. (Resp., Dkt. 18, at 20). Again, the member participation requirement is purely “prudential” and is not fatal to Plaintiffs' standing. See Assn. of Am. Physicians & Surgeons, Inc., 627 F.3d at 550. Unlike the laws at issue in Moody, HB 18 imposes relatively uniform requirements on all Plaintiffs' covered members. For example, if HB 18 is problematic because it requires DSPs to filter broad swaths of covered content, that problem applies equally across DSPs. See infra, Section III.C.1. The law's overbroad tailoring does not vary between covered DSPs. The same is true for vagueness-there is no reason to believe one social media company is better suited than another to understand HB 18's vague terms.
4. Ripeness for As-Applied Challenges
As to ripeness, Paxton suggests that “a party cannot challenge a statute as-applied unless the statute has been applied to him.” (Resp., Dkt. 18, at 21 (quoting Ctr. for Individual Freedom v. Carmouche, 449 F.3d 655, 659 (5th Cir. 2006))). But in the First Amendment context, a party does “not need to break the law and thereby expose [themselves] to liability” in order to challenge a regulation. Braidwood Mgmt., Inc. v. EEOC, 70 F.4th 914, 926 (5th Cir. 2023). The general test for ripeness, which the Fifth Circuit has used in the as-applied context, see id., examines only “(1) the fitness of the issues for judicial decision and (2) the hardship to the parties of withholding court consideration.” Book People, 91 F.4th at 333. That test is met. Legal issues predominate this challenge and are ready for judicial review, while withholding that review would impose irreparable harm on Plaintiffs' members. See infra, Section III.F.
5. Third-Party Standing
Last, Paxton argues that Plaintiffs lack the ability to assert the First Amendment rights of the users of social media DSPs. (Resp., Dkt. 18, at 21-22). As to the provisions which cover users' ability to post and view speech, Plaintiffs have third-party standing. Am. Booksellers, 484 U.S. at 392 (1988) (“[I]n the First Amendment context, litigants are permitted to challenge a statute not because their own rights of free expression are violated, but because of a judicial prediction or assumption that the statute's very existence may cause others not before the court to refrain from constitutionally protected speech or expression.”) (cleaned up). It is less clear that Plaintiffs have standing to assert the rights of users related to the other provisions, such as collection of geolocation data. Those provisions do not seem to regulate a user's access to speech, so it is unclear that the American Booksellers presumption would apply. Instead, Plaintiffs' standing for those provisions seems to be based on arguments that the law is facially overbroad and would burden the ability to host protected speech. (See Compl., Dkt. 1, at 2-3). Assuming those provisions do burden the ability to display protected speech, then third-party standing is unnecessary because social media DSPs have standing to sue in their own right.
C. First Amendment Analysis
Having found that Plaintiffs have standing to bring suit, the Cout turns its analysis to the merits of the case. “[T]he First Amendment provides in relevant part that ‘Congress shall make no law . . . abridging the freedom of speech,' and ‘the Fourteenth Amendment makes the First Amendment's Free Speech Clause applicable against the States.'” NetChoice, LLC v. Fitch, No. 1:24-CV-170-HSO-BWR, 2024 WL 3276409, at *7 (S.D.Miss. July 1, 2024) (quoting Manhattan Cmty. Access Corp. v. Halleck, 587 U.S. 802, 808 (2019)). The First Amendment protects both freedom of speech as well as the “right to receive information and ideas, regardless of their social worth.” Stanley v. Georgia, 394 U.S. 557, 564 (1969). First Amendment protection extends to social media, which “allows users to gain access to information and communicate with one another about it on any subject that might come to mind.” Packingham v. North Carolina, 582 U.S. 98, 107 (2017). “[T]o foreclose access to social media altogether is to prevent the user from engaging in the legitimate exercise of First Amendment rights.” Id. at 108.
1. Strict Scrutiny Applies
The threshold question is what level of scrutiny applies to HB 18's regulations. See Turner Broad. System, Inc. v. FCC, 512 U.S. 622, 637 (1994). In their motion, Plaintiffs treat HB 18 as a whole, arguing that the law's “central coverage definition of ‘digital service provider' is” a content-and speaker-based regulation. (Mot. Prelim. Inj., Dkt. 6, at 14). In response, Paxton suggests that it is a mistake to treat HB 18 as one indivisible law, subject to the same standard of scrutiny across the board. (Resp., Dkt. 18, at 23). Instead, Paxton argues that the Court must treat each subsection separately, conducting a “provision-by-provision” analysis rather than a “one-size-fits-all” approach. (Id. at 24).
The Court agrees with Plaintiffs that HB 18's threshold coverage definition is a contentbased regulation. HB 18 § 509.002(a)(1). HB 18 regulates [DSPs] that specifically host or broadcast “social” speech, thereby subjecting “social” content to heightened regulation. Id. HB 18 § 509.002(a)(1). Non-social interactions, such as professional interactions, are not covered, while social interactions are. That is clear in HB 18's content-based exceptions. DSPs that provide “content primarily generated or selected by the” provider or primarily “functions to provide a user with access to news, sports, [or] commerce” are exempted. HB 18 § 509.002(b)(10)(A). This “singles out specific subject matter for differential treatment,” by using the “function or purpose” of speech as a stand-in for its content. Reed v. Town of Gilbert, 576 U.S. 155, 163-64, 169 (2015). HB 18 is therefore a content- and speaker-based regulation, targeting DSPs whose primary function is to share and broadcast social speech. “[L]aws favoring some speakers over others demand strict scrutiny when the legislature's speaker preference reflects a content preference.” Id. at 170. When the government favors some speakers over others for their content, the law must be subject to strict scrutiny. Barr v. Am. Assn. of Political Consultants, Inc., 591 U.S. 610, 619-21 (2020) (controlling plurality op.); Reed, 576 U.S. at 163 (“Government regulation of speech is content based if a law applies to particular speech because of the topic discussed or the idea or message expressed.”). The elevation of news, sports, commerce, and provider-generated content over user-generated content is a content-based regulation.
In response, Paxton suggests that these arguments are foreclosed by the Fifth Circuit. (Resp., Dkt. 18, at 24 (citing NetChoice, LLC v. Paxton, 49 F.4th 439, 480 (5th Cir. 2022) (“NetChoice I”), vacated and remanded sub nom. Moody v. NetChoice, LLC, 144 S.Ct. 2383 (2024))). In NetChoice I, the Fifth Circuit rejected a similar argument brought by Plaintiffs by holding that regulations targeting social media did “not render [the law at issue] content-based because the excluded websites are fundamentally dissimilar mediums.” NetChoice I, 49 F.4th at 480.
That ruling is no longer binding because the Supreme Court vacated NetChoice I, “void[ing] each of the judgment's holdings.” Doe v. McKesson, 71 F.4th 278, 286 (5th Cir. 2023); see also Moody, 144 S.Ct. at 2409 (vacating judgment). Paxton suggests that Moody “effectively confirmed, or at least did not disturb, the Fifth Circuit's analysis on this point.” But the Supreme Court “disturbed” the analysis when it vacated the opinion. Paxton suggests that the Supreme Court's own opinion “also rebuffed” Plaintiffs theory-but it is not clear how. (See Resp., Dkt. 18, at 24). To the contrary, the Supreme Court expressly stated that “there has been enough litigation already to know that the Fifth Circuit, if it had stayed the course, would get wrong at least one significant input ....” Moody, 144 S.Ct. at 2349. While the Supreme Court did not determine “whether to apply strict or intermediate scrutiny[,]” that was only because “Texas's law [did] not pass” either intermediate or strict scrutiny, at least applied to key respects of the law. Id. at 2407.
Cutting against the Fifth Circuit's analysis in NetChoice I, the Supreme Court noted that “presenting a curated and edited compilation of [third-party] speech is itself protected.” Id. at 2409 (internal citations omitted). It noted that “editorial judgments influencing the content of [Facebook's News Feed and YouTube's homepage] are, contrary to the Fifth Circuit's view, protected expressive activity.” Id. If the discretion to favor social content over other messages is protected editorial judgment, then laws like HB 18 are not just targeting “fundamentally dissimilar mediums.” Those laws are targeting DSPs based on those protected editorial judgments. So, at the very least, Moody calls into serious question NetChoice I's discussion of “content-based” laws.
The district court's reasoning in NetChoice, LLC v. Yost is instructive. No. 2:24-CV-00047, 2024 WL 555904, at *8-11 (S.D. Ohio Feb. 12, 2024). There, a district court dealt with a highly similar Ohio law that targeted DSPs hosting “social” speech and excluded sites focused on news and product reviews. Id. at *8. The district court grappled with whether the “social” distinction was “content-based” or merely regulated the “manner in which” companies “transmit” the content. Id. at *9 (quoting Turner Broad. Sys., 512 U.S. at 637). Although emphasizing that “[i]t is a close call[,]” the court ultimately found that social media DSPs “are not ‘mere conduits'” but ‘curate both users and content to convey a message about the type of community the platform seeks to foster.'” Id. at *11 (quoting NetChoice, LLC v. Paxton, 573 F.Supp.3d 1092, 1107 (W.D. Tex. 2021)).
But see Reed, 576 U.S. at 163 (“Some facial distinctions based on a message are obvious, defining regulated speech by particular subject matter, and others are more subtle, defining regulated speech by its function or purpose. Both are distinctions drawn based on the message a speaker conveys, and, therefore, are subject to strict scrutiny.”).
In NetChoice I, the Fifth Circuit ruled that these editorial decisions fall outside the First Amendment, which in turn allowed the court to consider social media platforms “mere conduits.” NetChoice I, 49 F.4th at 480. But in reversing NetChoice I on the First Amendment value of editorial judgments, Moody undercut the premise that had allowed the Fifth Circuit to consider social media regulations content neutral. Moody, 144 S.Ct. at 2399.
In an even more similar case, a Mississippi district court found a nearly identical law (H.B. 1126) to be content- and speaker-based:
In essence, H.B. 1126 treats or classifies digital service providers differently based upon the nature of the material that is disseminated, whether it is “social interaction,” as opposed to “news, sports, commerce, [or] online video games[.]” [HB 1126] can thus be viewed as either drawing a facial distinction based on the message the digital service provider conveys (i.e., news and sports), or based on a more subtle content-based restriction defining regulated speech by its function or purpose (i.e., providing news and sports). Either way, [b]oth are distinctions drawn based on the message a speaker conveys.Fitch, 2024 WL 3276409, at *9 (quoting HB 1126) (cleaned up).
Like the district courts in Yost and Fitch, this Court finds that HB 18 discriminates based on the type of content provided on a medium, not just the type of medium. A DSP that allows users to socially interact with other users but “primarily functions to provide” access to news or commerce is unregulated. An identical DSP, with the exact same medium of communication and method of social interaction, but “primarily functions to provide” updates on what a user's friends and family are doing (e.g., through Instagram posts and stories), is regulated. If there is a difference between the regulated DSP and unregulated DSP, it is the content of the speech on the site, not the medium through which that speech is presented. When a site chooses not to primarily offer news but instead focus on social engagement, it changes from an uncovered to covered platform. But the type of medium has not changed, only the content primarily expressed on the platform.
In sum, strict scrutiny applies to HB 18's provisions because the law regulates DSPs based on the content of their speech and the identity of the speaker. Paxton must therefore prove that HB 18 is “the least restrictive means of achieving a compelling state interest.” Nat'l Inst. of Family & Life Advocs. v. Becerra, 585 U.S. 755, 766 (2018) (“NIFLA”) (citation omitted).
2. HB 18 and Strict Scrutiny
In his response, Paxton does not mount a defense on whether HB 18 passes strict scrutiny as a whole. (See Resp., Dkt. 18, at 25-28). Instead, Paxton goes “provision-by-provision,” arguing that some restrictions pass scrutiny and that even if others fail First Amendment scrutiny, they can be severed and do not mean the entire law is unconstitutional.
Here, the Court agrees with Paxton. Even if HB 18 is a content-based regulation, it does not follow as a matter of course that the law is facially invalid. In the First Amendment context, facial challenges can only succeed if litigants show that “a substantial number of [the law's] applications are unconstitutional, judged in relation to the statute's plainly legitimate sweep.” Ams. for Prosperity Found. v. Bonta, 594 U.S. 595, 615 (2021). So a law regulating First Amendment activity may only be struck down in its entirety if its “unconstitutional applications substantially outweigh its constitutional ones.” Moody, 144 S.Ct. at 2397.
Moody, including the majority opinion and all four concurrences, emphasized that courts should not treat facial challenges lightly, even in the First Amendment context. It clarified that courts should “address the full range of activities the laws cover, and measure the constitutional against the unconstitutional applications.” Id. at 2397-98. That analysis requires a two-step process. First, courts must “assess the state law's scope” and ask, “What activities, by what actors, do the laws prohibit or otherwise regulate?” Id. at 2398. Second, a court must “decide which of the laws' applications violate the First Amendment, and to measure them against the rest.” Id. Only after making these inquiries can a court determine if a law's “unconstitutional applications substantially outweigh its constitutional ones.” Id. at 2397.
It is a mistake to treat HB 18 as a whole and assume that because the law draws a threshold content-based distinction, it fails entirely under strict scrutiny. Per Moody, the Court must instead go through each provision and examine its respective permissibility under the First Amendment.
a. The Data Privacy, Parental Control, and Disclosure Provisions
The Court first “assess[es] the state law's scope.” Moody, 144 S.Ct. at 2398. While Plaintiffs focus their challenge on the monitoring-and-filtering requirements, HB 18 imposes various other requirements. (See Mot. Prelim. Inj., Dkt. 6). These other provisions, while subject to strict scrutiny, are largely unrelated to First Amendment expression. Section 509.052 prohibits covered DSPs from collecting unnecessary amounts of known minors' PII or geolocation data, allowing minors to make financial transactions through the DSP, sharing minors' PII, and targeting advertising to minors. HB 18 § 509.052. Another section requires covered DSPs to allow parents to control a minor's privacy settings and limit the amount of time a minor spends on the DSP. Id. § 509.054. Covered DSPs must also disclose an overview of their algorithms for providing content; promoting, ranking, and filtering content; and the PII used to provide content. Id. § 509.056.
Moody's next step is to “decide which of the laws' applications violate the First Amendment, and to measure them against the rest.” 144 S.Ct. at 2398. However, in their briefing, Plaintiffs do not clearly do so. It is Plaintiffs' burden to show that the law is unconstitutional in these respects. Indeed, for a preliminary injunction, Plaintiffs must make a “clear showing” that they are entitled to the relief they seek. Mazurek v. Armstrong, 520 U.S. 968, 972 (1997) (citing 11A C. Wright, A. Miller, & M. Kane, Federal Practice and Procedure § 2948, pp. 129-130 (2d ed. 1995)). That clear showing requires more than the law as whole to be “presumptively unconstitutional.” Reed, 576 U.S. at 163.
It remains possible that each provision will fail under strict scrutiny. But that is not a given. And it is not certain to be the case under HB 18, where many provisions seem to regulate conduct and only incidentally burden speech (if at all). See Moody, 144 S.Ct. at 2402 n.4. Plaintiffs do not show how Section 509.052 places any burden on speech by prohibiting the collection of PII and geolocation data. This is primarily a regulation of conduct, so it is not clear that the law restricts or even burdens speech. Similarly, it is not clear that a law requiring parents to be allowed to access and change their children's privacy settings implicates First Amendment concerns. Overall, these provisions likely primarily regulate conduct, and while the Court can conceive of ways in which they do burden speech (e.g., reducing the hours a child may spend consuming speech on social media), that point is not sufficiently developed at this stage.
In sum, Plaintiffs fail to make a “clear showing” that these provisions are unconstitutional. Those provisions, then, do not need to be enjoined, because it is not clear that, taken as a whole, “a substantial number of [those provisions'] applications are unconstitutional ....” Ams. for Prosperity Found., 594 U.S. at 615.
In addition, as Paxton notes, the remaining provisions are potentially severable, as the law does not necessarily operate as an indivisible whole.
b. The Monitoring-and-Filtering Requirements
While the preceding sections of HB 18 may be constitutional, the same cannot be said for the monitoring-and-filtering requirements. These requirements force providers to develop strategies to “prevent [a] known minor's exposure to harmful material and other content that promotes, glorifies, or facilitates: (1) suicide, self-harm, or eating disorders; (2) substance abuse; (3) stalking, bullying, or harassment; or (4) grooming, trafficking, child pornography, or other sexual exploitation or abuse.” HB 18 § 509.053. Irrespective of whether HB 18 as a whole is content-based, there can be little dispute that this provision is. The monitoring-and-filtering requirements explicitly identify discrete categories of speech and single them out to be filtered and blocked. That is as content based as it gets.
It is far from clear that Texas has a compelling interest in preventing minors' access to every single category of information listed above. Some interests are obvious-no reasonable person could dispute that the state has a compelling interest in preventing minors from accessing information that facilitates child pornography or sexual abuse. See Sable Commc'ns of California, Inc. v. FCC, 492 U.S. 115, 126 (1989) (“[T]here is a compelling interest in protecting the physical and psychological well-being of minors.”). On the other end, many interests are not compelling, such as regulating content that might advocate for the deregulation of drugs (potentially “promoting” “substance abuse”) or defending the morality of physician-assisted suicide (likely “promoting” “suicide”). See Brown v. Ent:. Merchants Assn, 564 U.S. 786, 794-95 (2011) (“No doubt a State possesses legitimate power to protect children from harm, but that does not include a free-floating power to restrict the ideas to which children may be exposed.”) (internal citation omitted). The Supreme Court has repeatedly emphasized that “[s]peech that is neither obscene as to youths nor subject to some other legitimate proscription cannot be suppressed solely to protect the young from ideas or images that a legislative body thinks unsuitable for them.” Erznoznik v. Jacksonville, 422 U.S. 205, 213-13 (1975). Much of the regulated topics are simply too vague to even tell if it is compelling. Terms like “promoting,” “glorifying,” “substance abuse,” “harassment,” and “grooming” are undefined, despite their potential wide breadth and politically charged nature. While these regulations may have some compelling applications, the categories are so exceedingly overbroad that such a showing is unlikely.
Even accepting that Texas has a compelling interest in blocking select categories under HB 18, the law is not narrowly tailored. HB 18 requires DSPs to implement strategies that include: listing harmful material, using filtering technology, using hash-sharing technology to identify recurring material, creating a database of keywords used for filter evasion, performing humanmonitoring reviews to ensure the filtering technology works, publishing descriptions of the filtered content, and making the code available to independent researchers. HB 18 § 509.053(b)(1). This broad, multi-faceted approach fails for two reasons: it does not employ the least restrictive means, and it is not narrowly tailored.
As in Fitch, Paxton “has not shown that the alternative suggested by [Plaintiffs], a regime of providing parents additional information or mechanisms needed to engage in active supervision over children's internet access would be insufficient to secure the State's objective of protecting children.” 2024 WL 3276409, at *12. By contrast, Plaintiffs have demonstrated that many DSPs do implement content-moderation policies to ensure that minors cannot access harmful content. (Mot. Prelim. Inj., Dkt. 6, at 22). And Paxton has not shown that methods such as “hash-sharing technology” and publishing depictions of filtered content are necessary to prevent harm to minors. In short, HB 18 does not employ “the least restrictive means” to stop minors from accessing harmful material. See United States v. Playboy Ent. Grp., 529 U.S. 803, 813 (2000).
HB 18 also employs overbroad terminology. Again, the monitoring-and-filtering requirements impose sweeping ex-ante speech restrictions, akin to prior restraints, but does little more than vaguely gesture at what speech must be restrained. For example, what does it mean for content to “promote” “grooming?” The law is not clear. So, by requiring filtering as a matter of law with only vague reference to what must be filtered, HB 18 will likely filter out far more material than needed to achieve Texas's goal.
Separate from their other arguments, Plaintiffs argue that the monitoring-and-filtering provisions are unconstitutional because they impose a prior restraint. (Mot. Prelim. Inj., Dkt. 6, at 18). Paxton responds that prior restraints must typically come from “administrative and judicial orders,” so HB 18 does not qualify. (Resp., Dkt. 18, at 44). Finding the provisions invalid under strict scrutiny (as well as vague and preempted), the Court does not decide the issue. There can be little question, though, that the monitoring-and-filtering provisions require DSPs to affirmatively seek out and block certain categories of speech, raising the same overbreadth concerns as with prior restraints. See FCC v. Fox Television Stations, Inc., 567 U.S. 239, 253 (2012) (“[L]aws which regulate persons or entities must give fair notice of conduct that is forbidden or required.”).
More problematically, the law is underinclusive. A law that “is wildly underinclusive when judged against its asserted justification . . . is alone enough to defeat it.” Brown, 564 U.S. at 802. Websites that “primarily” produce their own content are exempted, even if they host the same explicitly harmful content such as “promoting” “eating disorders” or “facilitating” “self-harm.” The most serious problem with HB 18's under-inclusivity is it threatens to censor social discussions of controversial topics. “[S]ocial media in particular” operates as one of “most important places . . . for the exchange of views ....” Packingham, 582 U.S. at 104. But HB 18 specifically cuts teenagers off from this critical “democratic forum[] of the Internet” even though the same harmful content is available elsewhere. Reno v. ACLU, 521 U.S. 844, 868 (1997). A teenager can read Peter Singer advocate for physician-assisted suicide in Practical Ethics on Google Books but cannot watch his lectures on YouTube or potentially even review the same book on Goodreads. In its attempt to block children from accessing harmful content, Texas also prohibits minors from participating in the democratic exchange of views online. Even accepting that Texas only wishes to prohibit the most harmful pieces of content, a state cannot pick and choose which categories of protected speech it wishes to block teenagers from discussing online. Brown, 564 U.S. at 794-95.
Of course, Moody's instruction on facial challenges still applies: the Court must analyze what each provision of the monitoring-and-filtering requirements does and individually assess those provisions' constitutionality. Here, however, the strict scrutiny test parallels the facial challenge framework. See United States v. Edge Broad. Co., 509 U.S. 418, 430 (1993) (“[T]he validity of [a] regulation depends on the relation it bears to the overall problem the government seeks to correct, not on the extent to which it furthers the government's interest in an individual case.”); NetChoice, LLC v. Bonta, No. 23-2969, 2024 WL 3838423, at *8 (9th Cir. Aug. 16, 2024) (slip op.) (affirming facial challenge where strict scrutiny regulation “raises the same First Amendment issues” “in every application to a covered business”); United States v. Stevens, 559 U.S. 460, 473 n.3 (2010) (noting that overbroad law was facially invalid).
The monitoring-and-filtering provisions are more restrictive and less tailored than they need to be at the outset. The provisions' facial under-inclusivity cannot be cured by identifying examples of HB 18's constitutional applications-the point is that HB 18 does not regulate the same harmful content elsewhere. The lack of narrow tailoring and use of overly restrictive means apply to all speech regulated under the provision and “raises the same First Amendment issues.” NetChoice, LLC v. Bonta, 2024 WL 3838423, at *8. Nor are the monitoring-and-filtering requirements like the laws at issue in Moody, where some provisions might restrict protected speech and others do not. 144 S.Ct. at 2398. Instead, the monitoring-and-filtering requirements exclusively target speech, only a small portion of which falls outside First Amendment coverage. Because the monitoring-and-filtering requirements are overbroad, overly restrictive, and underinclusive, they are properly enjoined on their face.
For example, child pornography generally does not receive First Amendment protection. See New York v. Ferber, 458 U.S. 747, 765 (1982). But the large majority of the proscribed content, such as speech promoting eating disorders, harassment, and bullying, is covered speech, even if highly distasteful. A state may not “create a wholly new category of content-based regulation that is permissible only for speech directed at children.” Brown, 564 U.S. at 794.
The preliminary injunction shall also apply to Section 509.056(1), which requires DSPs to “make a commercially reasonable effort to ensure that [their] algorithm does not interfere with the digital service provider's duties under” the monitoring-and-filtering requirements.
D. Vagueness
In the alternative to their First Amendment challenge, Plaintiffs argue that HB 18's definition for covered DSPs and its monitoring-and-filtering requirements are void for vagueness. (Mot. Prelim. Inj., Dkt. 6, at 24-25). As a reminder, HB 18 covers DSPs that “allow[] users to socially interact with other users” but exempts DSPs that “primarily function[] to provider a user with access to news, sports, commerce, or content primarily generated or selected by the DSP ....” HB 18 §§ 509.002(a), (b)(10)(A). Plaintiffs take issue with the terms “socially interact” and “primarily functions” as they define covered DSPs. (Id. at 17). And as to the monitoring-and-filtering requirements, Plaintiffs take issue with the terms “promotes,” “glorifies,” and “facilitates,” as well as the categories of proscribed topics. (Id. at 24-25).
Plaintiffs also argue that the law is vague because it exempts services that only have “incidental” social functions but does not define “incidental.” (Mot. Prelim. Inj., Dkt. 6, at 17). Plaintiffs do not allege, however, that any members are genuinely on the threshold where this ambiguity might come into effect. And such terminology has been upheld by a previous Supreme Court decision. McGowan v. State of Md., 366 U.S. 420, 428-29 (1961).
“A fundamental principle in our legal system is that laws which regulate persons or entities must give fair notice of conduct that is forbidden or required.” FCC v. Fox Television Stations, Inc., 567 U.S. 239, 253 (2012). “A law is unconstitutionally vague if it (1) fails to provide those targeted by the statute a reasonable opportunity to know what conduct is prohibited, or (2) is so indefinite that it allows arbitrary and discriminatory enforcement.” McClelland v. Katy Indep. Sch. Dist., 63 F.4th 996, 1013 (5th Cir.), cert. denied 144 S.Ct. 348 (2023). “A regulation is void for vagueness when it is so unclear that people ‘of common intelligence must necessarily guess at its meaning and differ as to its application.'” Id. (quoting Connally v. Gen. Constr. Co., 269 U.S. 385, 391 (1926)). The degree of vagueness that the Constitution tolerates-as well as the relative importance of fair notice and fair enforcement-depends in part on the nature of the enactment.” Vill. of Hoffman Ests. v. Flipside, Hoffman Ests., Inc., 455 U.S. 489, 498 (1982). The Supreme Court has “expressed greater tolerance of enactments with civil rather than criminal penalties because the consequences of imprecision are qualitatively less severe.” Id. at 498-99. However, if “the law interferes with the right of free speech or association, a more stringent vagueness test should apply.” Id. at 499.
1. The Covered Entity Definitions
Starting with the covered entity definitions, the Court is not convinced that the terms are so vague as to render the entire law unconstitutional. Plaintiffs note that “it is unclear what test one uses to determine how a digital service ‘primarily' functions.” (Mot. Prelim. Inj., Dkt. 6, at 17 (citing Fitch, 2024 WL 3276409, at *15)). That argument is a stretch. “Primary” has a fairly ascertainable meaning: “for the most part” or “in the first place.” Primarily, Webster's New World Dictionary (4th ed. 2013). When HB 18 exempts DSPs that “primarily function[] to provide a user with access to news, sports, commerce,” it likely targets DSPs whose main function is one of those categories. Turning to the word “socially,” the Court also finds it is sufficiently definite to survive a preliminary injunction motion. The term is cabined within the definition of a DSP that “connects users in a manner that allows users to socially interact with other users on the digital service[.]” HB 18 § 509.002(1). Other regulations include allowing a user to create a public or semi-public profile or post content viewed by others. Id. §§ 509.002(2)-(3). Given the ample context surrounding the type of DSP covered, the term “socially interact” is capable of a limited common-sense meaning. See United States v. Comer, 5 F.4th 535, 542 (4th Cir. 2021) (denying vagueness challenge to law that targeted social networking).
Tellingly, there seems to be little (or no) actual confusion about whether Plaintiffs' members are covered under these allegedly vague terms. Facebook, Snapchat, X, and YouTube all appear to readily meet the statute's definition because they “primarily” offer “social interactions.” It remains possible that the laws are vague, and some providers, especially smaller DSPs, do not know whether they are covered. But such challenges would be more appropriate on an as-applied basis or after a genuine showing that a party does not know whether the law applies to them. See United States v. Wurzbach, 280 U.S. 396, (1930) (“[I]f there is any difficulty, which we are far from intimating, it will be time enough to consider it when raised by someone whom it concerns.”) (Holmes, J.). As it is, the covered entity definition “is not unconstitutionally vague” solely because it “engender[s] some limited amount of confusion ....” United States v. Clinical Leasing Serv., Inc., 925 F.2d 120, 123 (5th Cir. 1991).
2. The Monitoring-and-Filtering Provisions
The same cannot be said for the terminology used in HB 18's monitoring-and-filtering provisions. Again, those provisions require social media DSPs to track, filter, and block material that “promotes,” “glorifies,” or “facilitates” “suicide, self-harm,[] eating disorders[,] substance abuse[,] stalking, bullying, [] harassment[,] grooming, trafficking, child pornography,[] other sexual exploitation or abuse” and material that is sexually explicit for minors. HB 18 § 509.053. Those provisions are vague because both the verbs (promotes, glorifies, and facilitates) and the objects of those verbs (e.g., stalking, bullying, substance abuse, and grooming) are broad and undefined. Especially when put together, the provisions are unconstitutionally vague.
Begin with the verbs: promote, glorify, and facilitate. One of those words-“promote”-has already been held to be vague when regulating First Amendment activity. In Baggett v. Bullitt, 377 U.S. 360, 371-72 (1964), the Supreme Court dealt with a regulation that imposed a loyalty oath for teachers to swear that they will “promote respect for the flag and the institutions of the United States.” (emphasis added). The Supreme Court found that the term “promote” was “very wide indeed” and failed to “provide[] an ascertainable standard of conduct.” Id. In response, Paxton suggests that Baggett dealt with a “‘wildly different situation' than this one.” (Resp., Dkt. 18, at 38). But, if anything, the vagueness is more problematic under HB 18, because the law requires social media DSPs to guess which broad categories of speech, likely constituting billions of posts, must be filtered from view. So the wide-ranging meanings of “promote” will result in wide-ranging censorship of speech.
The problem is even more acute with the term “glorifying.” The word encompasses so wide an ambit that people “of common intelligence” can do no more than guess at its application. McClelland, 63 F.4th at 1013. To “glorify” potentially includes any content that favorably depicts a prohibited topic, leaving no clear answer on what content must be filtered. Do liquor and beer advertisements “glorify” “substance abuse?” Does Othello “glorify” “suicide?” Given the substantial liability companies face for failing to comply (to say nothing of the private rights of action), it is reasonable to expect that companies will adopt broad definitions that do encompass such plainly protected speech.
In sum, HB 18 imposes requirements for entire categories of speech to be blocked. If those ex-ante restrictions are to be imposed, they cannot arguably cover such broad categories of protected speech. Grayned v. City of Rockford, 408 U.S. 104, 109 (1972) (“[W]here a vague statute abuts upon sensitive areas of basic First Amendment freedoms, it operates to inhibit the exercise of (those) freedoms. Uncertain meanings inevitably lead citizens to steer far wider of the unlawful zone than if the boundaries of the forbidden areas were clearly marked.”) (cleaned up).
Paxton contends that the monitoring-and-filtering provisions are not vague to Plaintiffs because they come “directly from their own content-moderation policies.” (Resp., Dkt. 18, at 12 (emphasis omitted)). But a selfpoliced rule does not suffer the same vagueness problems as a state-backed proscription. Facebook may know their internal definition of “glorify” but the company cannot be assured that Paxton or Texas courts will abide by the same definition.
Paxton suggests that these terms, especially “facilitate,” can be limited to speech that helps bring about an unlawful act, is intended to induce illegal activities, or provides assistance to a wrongdoer. (Resp., Dkt. 18, at 23-24 (quoting Williams, 553 U.S. at 290, then quoting United States v. Hansen, 599 U.S. 762, 770-73 (2023))). In other words, Paxton suggests that the terms can be limited to “speech integral to unlawful conduct.” Id. But if the state had intended to proscribe only speech “integral to unlawful conduct,” it could have explicitly stated so. Instead, the state used three terms: “promote, glorify, and facilitate,” giving a broader context than criminal assistance. “Glorify” and “promote” taint the understanding of “facilitate” beyond direct criminal advocacy. And there are good reasons to believe the state deliberately sought a broader definition, because HB 18 also proscribes speech related to lawful conduct, such as having an eating disorder. Paxton's own introduction shows that many of social media's most deleterious effects comes from the incessant, favorable depiction of certain content, rather than its direct advocacy. (See Resp., Dkt. 18, at 13-16). As a result, HB 18 is not “readily susceptible” to the “narrowing construction” that Paxton suggests. Am. Booksellers, 484 U.S. at 397 (“[T]he statute must be readily susceptible to the limitation; we will not rewrite a state law to conform it to constitutional requirements.”) (internal quotation omitted).
The final issue for HB 18 is that the law fails to define key categories of prohibited topics, including “grooming,” “harassment,” and “substance abuse.” At what point, for example, does alcohol use become “substance abuse?” When does an extreme diet cross the line into an “eating disorder?” What defines “grooming” and “harassment?” Under these indefinite meanings, it is easy to see how an attorney general could arbitrarily discriminate in his enforcement of the law. See Smith v. Goguen, 415 U.S. 566, 575 (“Statutory language of such a standardless sweep allows [] prosecutors[] and juries to pursue their personal predilections.”). These fears are not too distant-pro-LGBTQ content might be especially targeted for “grooming.” See Little v. Llano Cnty., No. 1:22-CV-424-RP, 2023 WL 2731089, at *2 (W.D. Tex. Mar. 30, 2023) (finding that several books supporting pro-LGBTQ views were removed from library shelves for allegedly promoting “grooming”), aff'd as modified, 103 F.4th 1140 (5th Cir. 2024), reh'g en banc granted, opinion vacated, 106 F.4th 426 (5th Cir. 2024). Content related to marijuana use might be prosecuted as “glorifying” “substance abuse,” even if cigarette and alcohol use is not. This vast indefinite scope of enforcement would “effectively grant[] [the State] the discretion to [assign liability] selectively on the basis of the content of the speech.” City of Houston, Tex. v. Hill, 482 U.S. 451, 465 n.15 (1987). Such a sweeping grant of censorial power cannot pass First Amendment scrutiny.
E. Section 230 Preemption
Finally, Plaintiffs seek a preliminary injunction against the “monitoring-and-filtering” requirements by arguing that they are preempted by Section 230. Section 230 states that “[n]o interactive computer service shall be treated as the publisher or speaker of any information provided by” someone else. 47 U.S.C. § 230(c)(1). Under Section 230, Congress provided all DSPs “broad immunity” for “all claims stemming from their publication of information created by third parties.” Free Speech Coal., Inc. v. Paxton, 95 F.4th 263, 286 (5th Cir. 2024) (cleaned up), cert. granted 2024 WL 3259690 (Mem.) (U.S. July 2, 2024). That includes purported “failure[s] to implement basic safety measures to protect minors.” Doe v. MySpace, Inc., 528 F.3d 413, 418-19 (5th Cir. 2008). By preempting “inconsistent” laws, 47 U.S.C. § 230(e)(3), Congress barred regulation of websites' “decisions relating to the monitoring, screening, and deletion of content[.]” Id. at 420 (citation omitted).
HB 18 requires DSPs to “implement a strategy to prevent [a] known minor's exposure to” certain third-party content. HB 18 § 509.053(a). That requirement impermissibly forces covered DSPs to “address certain harmful content” on their services. MySpace, 528 F.3d at 420 (quoting Green v. Am. Online (AOL), 318 F.3d 465, 471 (3d Cir. 2003)). Similarly, HB 18 requires DSPs to monitor and filter certain categories of content. HB 18 § 509.053. But Section 230 “specifically proscribes liability” for “decisions relating to the monitoring, screening, and deletion of content.” Id.; La'Tiejira v. Facebook, Inc., 272 F.Supp.3d 981, 993 (S.D. Tex. 2017) (“Thus ‘any activity that can be boiled down to deciding whether to exclude material that third parties seek to post online is perforce immune under section 230.'” (quoting Fair Hous. Council of San Fernando Valley v. Roommates.com, LLC, 521 F.3d 1157, 1170 (9th Cir. 2009)). “[T]hat is the point of Section 230: to immunize [websites] for harm caused by unremoved speech.” Free Speech Coal., 95 F.4th at 284-85.
Paxton, in response, relies on the Fifth Circuit's recent decision in Free Speech Coalition. (Resp., Dkt. 18, at 43 (citing Free Speech Coal. 95 F4th at 284)). In that case, a group of websites hosting adult videos sued to enjoin a Texas law, HB 1181, that required the websites to implement age verification measures to screen out minors. Id. at 267. The Fifth Circuit determined that HB 1181 was “not reliant on the harm done by third-party content” because the law “impose[d] liability purely based on whether plaintiffs comply with the statute, independently of whether the third-party speech that plaintiffs host harms anybody.” Id. at 298. Here, Paxton draws a similar analogy, arguing that HB 18 imposes liability only for failing to comply with the statute, not for hosting content.
HB 18's monitoring and filtering requirements, unlike HB 1181, do impose liability based on the type of content that a website hosts. A website only filters out content that falls into certain categories of prohibited speech (e.g., “glorifying” an “eating disorder” and “promoting” “substance abuse”). The monitoring-and-filtering requirements necessarily derive their liability from the type of content a site displays. HB 18 § 509.151-152.
Imagine that Texas passed a law stating, “Social media websites must remove defamatory content.” Under Paxton's broad reading of Free Speech Coalition, the law would not be preempted because liability attaches based on whether a website complies with the law, not based on its content. That reasoning would altogether nullify Section 230 by having the same effect as directly imposing liability on the website for hosting third-party content. Section 230 provides “broad immunity” for providers for “all claims stemming from their publication of information created by third parties.” MySpace, 528 F.3d at 418 (emphasis added). Liability under HB 18 stems from the content it hosts, even if liability directly attaches based on compliance with the law. Accordingly, the Court finds that Section 230 preempts HB 18's monitoring and filtering requirements.
F. Remaining Preliminary Injunction Factors
Paxton does not seriously contest the remining preliminary injunction factors. (Resp., Dkt. 18, at 45). He notes that Plaintiffs waited more than a year to bring suit, cutting against their argument for irreparable harm. (Id.). Lengthy delays in filing suit do count against irreparable harm, especially in the context of a temporary restraining order (“TRO”). See Luckenbach Texas, Inc. v. Skloss, No. 1:21-CV-871-RP, 2022 WL 5568437, at *4-5 (W.D. Tex. July 8, 2022); Embarcadero Techs., Inc. v. Redgate Software, Inc., No. 1:17-CV-444-RP, 2017 WL 5588190, at *3 (W.D. Tex. Nov. 20, 2017). That doctrine is less applicable here, where the law created a 14-month gap between enactment and taking effect. While the Court would have appreciated more time to decide these complex issues, Plaintiffs' delay does not automatically negate a finding of irreparable harm.
Setting aside their delay, Plaintiffs make a clear showing that they will suffer irreparable injury in the absence of an injunction. “The loss of First Amendment freedoms, for even minimal periods of time, unquestionably constitutes irreparable injury.” Roman Cath. Diocese of Brooklyn v. Cuomo, 592 U.S. 14, 19 (2020) (citation omitted). HB 18 will also impose nonrecoverable compliance costs, adding to Plaintiffs' irreparable harm. Rest. L. Ctr. v. U.S. Dept. of Lab., 66 F.4th 593, 597 (5th Cir. 2023) (“Under our precedent, the nonrecoverable costs of complying with a putatively invalid regulation typically constitute irreparable harm.”); Wages & White Lion Invs., L.L.C. v. United States Food & Drug Admin., 16 F.4th 1130, 1142 (5th Cir. 2021) (noting that sovereign immunity would bar the recovery of compliance costs). Declarations submitted by Plaintiffs highlight that many DSPs will be forced to substantially revamp their existing filtering software or develop it anew. (Mot. Prelim. Inj., Dkt. 6, at 27 (citing declarations)). Those compliance costs are greater because, as the Court has noted, Plaintiffs' members have no guidance on how to determine if content falls under HB 18's broad regulations. (Veitch Decl., Dkt. 6-4, at 19 (noting that HB 18 “will require alterations to YouTube's content moderation-not because YouTube's content moderation is currently deficient, but because of uncertainty of how the Texas Attorney General will interpret the Act.”)). For small DSPs that do not qualify as exempt small businesses, the content filtering requirements will be “prohibitive for a sole proprietor . . . to implement” leading to the DSPs potentially “shutting down [] rather than shoulder the[] compliance costs or run the risk of potential liability.” (Riley Decl., Dkt. 6-6, at 5-6). In sum, Plaintiffs have shown irreparable harm.
The balance of equities and public interest follow likelihood of success. These last two factors merge when the government is the opposing party. Nken v. Holder, 556 U.S. 418, 435 (2009). “[I]njunctions protecting First Amendment freedoms are always in the public interest.” Texans for Free Enter. v. Texas Ethics Comm'n, 732 F.3d 535, 539 (5th Cir. 2013). Because the Court has found that Plaintiffs have shown a substantial likelihood of success on the merits of their First Amendment claims, the Court finds that an injunction is in the public interest. See id.; see also, e.g., Book People, Inc., 91 F.4th at 341 (“Because Plaintiffs are likely to succeed on the merits of their First Amendment claim, the State and the public won't be injured by an injunction of a statute that likely violates the First Amendment.”).
IV. CONCLUSION
Plaintiffs have shown that HB 18 is a content-based statute and is therefore subject to strict scrutiny. As to the monitoring-and-filtering requirements, HB 18 § 509.053, Plaintiffs have carried their burden in showing that the law's restrictions on speech fail strict scrutiny, are unconstitutionally vague, and are preempted by Section 230. Because Plaintiffs also show that the remaining equitable factors weigh in their favor, the Court preliminarily enjoins Paxton from enforcing the monitoring-and-filtering provisions.
However, Plaintiffs do not show that the remaining provisions of HB 18 unconstitutionally regulate a meaningful amount of constitutionally protected speech or otherwise fail strict scrutiny. Accordingly, the preliminary injunction is limited only to HB 18's monitoring-and-filtering requirements, HB 18 §§ 509.053, 509.056(1).
Plaintiffs are not required to post a security bond. See Fitch, 2024 WL 3276409, at *17.
IT IS THEREFORE ORDERED that Plaintiffs' Motion for a Preliminary Injunction, (Dkt. 6), is GRANTED IN PART as set forth in this order. Plaintiffs' request in the alternative for a temporary restraining order is DISMISSED AS MOOT. Paxton's Motion for Leave to File a Surreply, (Dkt. 23), is DENIED.
IT IS FURTHER ORDERED that Paxton, his agents, employees, and persons working under his direction or control are PRELIMINARILY ENJOINED from enforcing the “monitoring-and-filtering” requirements, HB 18 §§ 509.053, 509.056(1), against Plaintiffs and their members pending a final judgment in this case or other modification of this order.
SIGNED.