First Amendment Notes 139 Harv. L. Rev. 942

Content Neutrality for Kids: Intermediate Scrutiny for Social Media Age-Verification Laws

Download
See Footnotes

The kids are not okay. Evidence of decreasing school performance,1 increasing rates of depression and anxiety, and declining social engagement among minors has created a national panic.2 One common culprit has emerged: social media companies.3 The Children’s Online Privacy Protection Act of 19984 is the current federal baseline for internet services to protect minors online, but it increasingly stretches credulity to accept social media companies’ assertion that they comply with the law.5 States have instead taken it upon themselves to enforce age restrictions on social media.6 Today, at least sixteen states have enacted laws regulating minors’ access to social media platforms.7

Social media age-verification laws8 raise immense First Amendment concerns: From data privacy issues, to chilling effects on social media posts, to potentially high costs for compliance,9 these laws “threaten[] to torch a large segment of the Internet community.”10 Companies have consequently brought successful First Amendment challenges in lower courts, blocking several attempts at regulation.11 And the Supreme Court is watching these developments closely; in a recent emergency docket concurrence, Justice Kavanaugh described Mississippi’s age-verification law as “likely unconstitutional.”12 So far, states have been mostly unsuccessful in defending their laws: All but one district court addressing the issue has held that these laws likely violate the First Amendment.

One of “the myriad tripwires laid for technology laws by current First Amendment doctrine” is strict scrutiny, the most rigorous form of judicial review.13 Given that strict scrutiny is “‘strict’ in theory and fatal in fact,”14 content-based regulations that merit strict scrutiny are all but certain to be unconstitutional. States can pass content-neutral regulations, which receive intermediate scrutiny. Intermediate scrutiny requires only that a law “advances important governmental interests unrelated to the suppression of free speech and does not burden substantially more speech than necessary to further those interests.”15 While this is still a significant hurdle (and one that many laws fail to clear), it gives states a fighting chance.

All but two cases evaluating these laws have held that their coverage provisions are content based given their targeting of “social” content or their use of content-based exemptions. Yet repeated losses in court haven’t stopped states from continuing to try,16 and other countries have implemented their own social media age-verification laws.17 While concrete evidence of a causal link between social media use and minors’ well-being remains contestable,18 concerns persist about social media’s propensity to cause addictive behaviors, mental health problems, and other harmful effects.19 The Supreme Court itself recently acknowledged that “social media pose[s] dangers . . . on adolescents’ mental health.”20 Many of these concerns will be exacerbated by the increased presence of artificial intelligence in social media.21

Thus, states seeking to enact social media age-verification laws must determine how to define social media in a way that targets platforms responsible for the alleged harms to minors without using content-based exemptions or covering platforms that are not responsible for these harms.

This Note argues that states can do so. It synthesizes lessons from lower court opinions addressing age-verification laws and recent Supreme Court precedent.22 Together, the cases indicate that states may be able to craft content-neutral laws by (1) focusing on specific, harmful features of social media, (2) refraining from using content-based exemptions, or (3) singling out harmful platforms under the reasoning of TikTok Inc. v. Garland.23 Part I discusses the components of First Amendment doctrine implicated by social media age-verification laws. Part II details the elements of these laws and the court opinions applying First Amendment scrutiny. Part III presents possible cures to the constitutional defects raised by the court opinions.

I.  Free Speech Doctrine Implicated by Social Media Age-Verification Laws

The First Amendment prohibits states and Congress from enacting laws “abridging the freedom of speech.”24 Both outright bans and indirect burdens on speech require heightened scrutiny under the First Amendment.25 Social media age-verification laws raise First Amendment concerns given their impacts on minors, parents, social media companies, and adults who post or engage with content on social media platforms. This Part describes the parts of free speech doctrine implicated by these laws.

A.  Content-Based Classifications

The First Amendment requires strict scrutiny for laws “that suppress, disadvantage, or impose differential burdens” based on the content of the regulated speech.26 Such laws “are presumptively invalid”27 and violate the fundamental norms of the First Amendment.28 The presumptive unconstitutionality of content-based regulations is rooted in the notion that individuals, rather than the government, should decide which ideas are worth engaging with.29 Prior to 2015, the key inquiry to determine whether a law was content based was if “the government . . . adopted [the] regulation of speech because of disagreement with the message it convey[ed].”30 Sorrell v. IMS Health Inc.31 illustrated the Court’s treatment of a content-based commercial regulation under the First Amendment: The Court held that a state law limiting the use of medical data for marketing purposes imposed a content-based burden on speech that merited heightened scrutiny.32 The Court also noted that laws that “burden[] disfavored speech by disfavored speakers” are similarly content based.33

Yet distinctions based on specific features of a medium of communication are not necessarily content based. In Turner Broadcasting System, Inc. v. FCC,34 the Supreme Court held that a federal law requiring cable television systems to carry local broadcast stations was content neutral.35 The challenged federal law distinguished between cable systems and broadcast stations, but the Court found that it did not preference broadcast stations based on the content of their programming.36 The Court also noted that the targeting of cable programs with subscribers above a set number was “justified by special characteristics of . . . cable medi[a],” notably its “monopoly power . . . and the dangers this power pose[d].”37 Members of the Turner Court38 and scholars have claimed that Turner’s reasoning was hard to square with the realities of the law,39 but the Court’s recent invocation of Turner suggests that Justices may be willing to rely on it for internet regulation.40

The Supreme Court clarified the standards for content-based regulations in two recent cases. First, in Reed v. Town of Gilbert,41 the Court stated that any regulation that “applies to particular speech because of the topic discussed or the idea or message expressed” is facially content based and warrants strict scrutiny.42 Laws that favor speakers are similarly content based when they “reflect[] a content preference.”43 Reed’s “bright-line test of content-based lawmaking” shifted the standard away from the previous purpose-oriented approach to the content-based inquiry.44 But in City of Austin v. Reagan National Advertising of Austin, LLC,45 the Court narrowed Reed by holding that a regulation that requires determining the “function or purpose” of speech “is [not] always content based.”46 Now, the critical inquiry is whether a regulation “single[s] out any topic or subject matter for differential treatment.”47

The difference between Reed and City of Austin is small, and the precise line between content-based and content-neutral regulations re­mains uncertain. This ambiguity is exacerbated by the lack of a clear definition for what constitutes “content.”48 Courts addressing social media age-verification laws have acknowledged the uncertainty in identifying when a regulation becomes content based.49 Nevertheless, post–City of Austin, laws may avoid strict scrutiny if they do not proscribe specific topics or messages.50

B.  Free Speech and the Internet

The First Amendment applies to internet speech. Though courts and scholars question the scope of First Amendment application to the internet, both social media users and platforms have free speech rights.51 The earliest attempts to regulate access to the internet concerned its propensity to provide access to adult content.52 In the late 1990s, Congress passed laws criminalizing online transmission or posting of content it deemed harmful to minors.53 In Reno v. ACLU54 and Ashcroft v. ACLU55 (Ashcroft II), the Court concluded these laws were content based because they sought to protect children from categories of speech protected by the First Amendment.56

Individuals have a First Amendment right to access social media platforms. In Packingham v. North Carolina,57 the Supreme Court struck down a statute58 barring registered sex offenders from accessing social media platforms used by minors.59 The Court described the internet as “the modern public square”60 and analogized online platforms to public parks to conclude that all adults had the right to access these platforms.61 Though Packingham’s public square analogy has been criticized extensively,62 it is generally accepted as the legal backdrop for online free speech rights.63 The Packingham Court did not preclude the possibility of constitutional internet regulation, but it made it clear that laws implicating internet speech would undergo heightened constitutional scrutiny.64

Social media companies have curation rights under the First Amendment. In Moody v. NetChoice, LLC,65 the Court analogized platforms to newspaper “editors, cable operators, and parade organizers,” all of whom engage in protected expressive activity in curating their information.66 Moody’s approach to internet regulation maintained the doctrinal status quo limiting the capacity of states to regulate the internet.67 The Court also made it clear that its conclusions did not depend on “the relative novelty of the technology.”68 Moody did not address whether the state laws in question were content based or content neutral, leaving the question open regarding what content neutrality looks like on the internet.69

C.  The 2024 Supreme Court Term

Two cases from the 2024 October Term suggest the Court may be increasingly open to certain forms of internet regulation. In TikTok and Free Speech Coalition, Inc. v. Paxton,70 the Court upheld two seemingly content-based internet regulations under intermediate scrutiny.71 While both cases are limited to their specific contexts — foreign corporate control of a social media platform72 and adult content “that is obscene for minors”73 — they provide insight into how the Court may approach future social media regulations. Both cases may portend a possible détente in “the Lochner era” of the First Amendment with respect to the internet,74 opening narrow yet viable avenues for states to enact social media age-verification laws.

In TikTok, the Court upheld a federal statute instituting a nationwide ban of the social media platform TikTok unless TikTok’s parent company, ByteDance, divested from its ownership in TikTok due to ByteDance’s potential control by the Chinese government.75 The Court concluded that the law was content neutral because it did “not target . . . speech based upon its content, . . . function[,] or purpose,”76 and held the law “satisfie[d] intermediate scrutiny.”77 In addressing the law’s targeting of TikTok, the Court quoted Turner to assert that a law is not content based when its singling out of a speaker reflects a “special characteristic”78 rather than “a content preference.”79

TikTok’s reliance on Turner suggests that regulations distinguishing between online speakers are not necessarily content based. The TikTok Court noted that the statute did not “impose a ‘restriction, penalty, or burden’ by reason of content,” and that TikTok could not comply with the law “by altering [its] speech.”80 While not directly overturning Reed, TikTok applied a more forgiving standard for determining whether a law is content based.81 The Court’s reliance on Turner also indicated a willingness to regulate social media companies on the basis of their “special characteristic[s],”82 a legal hook which may prove critical for defending social media age-verification laws.

In Paxton, the Court addressed states’ ability to limit minors’ access to sexual content.83 Minors have a well-established free speech right that includes “the right to receive information and ideas,”84 but this right is not as extensive as that of adults.85 In Ginsberg v. New York,86 the Court upheld an age-based limitation on minors’ access to sexual content,87 but in a later case it emphasized that states “could [not] create new categories of unprotected speech” beyond those historically recognized.88 In Paxton, the Court upheld a Texas law requiring age verification to access adult websites on the internet.89 Despite clear content-based distinctions,90 the Court concluded that the law merited only intermediate scrutiny since it “ha[d] ‘only an incidental effect on’” adults’ protected “First Amendment right to access speech that is obscene to minors.”91 Relying on Ginsberg, the Court held that Texas’s law lawfully targeted minors’ access to content that they do not have a First Amendment right to view.92

TikTok and Paxton suggest a possible shift in the longstanding advantage of companies challenging laws implicating speech.93 Given the narrower free speech rights of minors, there is greater latitude for social media age-verification laws if they avoid content-based designations disfavored in First Amendment precedent.

II.  Social Media Age-Verification Laws and Courts’ Scrutiny of Them Under the First Amendment

As of October 2025, at least sixteen states have passed laws that address minors’ use of social media.94 Internet companies have challenged the laws for violating the First Amendment, and the district courts that have addressed these laws have almost uniformly enjoined them for violating the First Amendment.95 This Part describes social media age-verification laws and how courts have approached the question of whether the laws are content based.96

A.  The Elements of Social Media Age-Verification Laws

While they differ in their approaches, social media age-verification laws generally define covered social media platforms, establish age-based restrictions on access, and impose penalties for companies that violate the law.

1.  Defining Covered Social Media Platforms. — Social media age-verification laws begin with coverage provisions that establish which platforms are covered. Utah’s law defines covered social media services as “public website[s] or application[s] that” display user-generated content; permit users to create accounts and profiles; allow users to create connections with other users and socially interact; and allow users to post publicly visible content.97 Other laws, such as those of Ohio and California, define covered platforms as those “that target[] children, or [are] reasonably anticipated to be accessed by children.”98

Many states’ laws make exceptions for certain websites or applications. Georgia’s law, for example, includes a “voluminous” list of exemptions99 that “create[s] a complicated web of what is (and is not) subject to the law.”100 These exemptions are often for platforms that cover primarily “news, sports,” professional services, “online video games,” and other types of content.101 Some states exempt platforms on “which interactions . . . are limited to commercial transactions or to consumer reviews.”102 And some states exempt platforms that “primarily generate[] or select[]” their own content.103

2.  Requirements Imposed on Social Media Companies. — Social media age-verification laws impose an array of different requirements on social media platforms to restrict the access of minors. Most involve some form of direct age verification or a less stringent age assurance.104 For example, Arkansas’s law requires the use of a third-party service to conduct age verification before an individual can access a social media platform.105 Utah’s law requires the use of an age assurance method that correctly estimates whether a user is a minor at least ninety-five percent of the time.106 A few of the laws do not actually mandate age verification or age assurance requirements,107 but challengers to these laws have argued they will still need to employ age verification to comply.108 Some laws place blanket prohibitions on creating social media accounts,109 while others restrict access to platforms entirely.110 Other laws restrict or limit minors’ access to only specific social media features.111 Enforcement is generally through civil penalties112 or a private right of action provided to parents or guardians.113

Many age-verification laws give parents control over whether a state law’s requirement applies to a minor. For example, Ohio’s law requires a platform to deny access to minors under sixteen114 unless it “[o]btain[s] verifiable consent . . . from the [minor]’s parent or legal guardian.”115 Similarly, California allows minors to bypass its prohibitions on access to algorithmic feeds116 and establishment of default platform settings with affirmative parental consent.117 Texas’s law empowers parents and guardians to “dispute[] the registered age of [a] minor”118 if they can verify their relationship with the minor.119 States that employ parental-control options argue that parents are best situated to determine whether their children should access social media.120

B.  The Content-Based Nature of Social Media Age-Verification Laws

Except for two cases, every court’s analysis of First Amendment challenges to social media age-verification laws has held that the laws are content based and has applied strict scrutiny.121 While each court has done so based on each law’s attempt to define covered platforms, the content-based inquiry is “not that simple” and it is uncertain where the line between content-based and content-neutral regulations of social media lies.122

1.  “Social” as a Content-Based Distinction. — Several courts have held that the designation of “social” content itself is a content-based distinction. For example, the Southern District of Ohio held that the designation of the features targeted by Ohio’s law — “functionalities allowing users to post, comment, and privately chat” — was a content-based distinction given its targeting of social content.123 The District of Utah similarly described Utah’s coverage provision as “divid[ing] the universe of internet platforms into social media services . . . that ‘allow users to interact socially with each other,’ and other internet platforms,” and held that the law “single[d] out . . . the ‘social’ subject matter” of social media.124 A number of other courts made similar holdings under Reed’s standard for facially content-based designations.125

A few courts, however, held that laws targeting social interactions do not make content-based distinctions. In Computer & Communications Industries Ass’n v. Uthmeier,126 the Northern District of Florida held that “‘social’ speech” was not “a topic or subject in the relevant sense” and thus not a category under City of Austin since social speech can cover “any conceivable topic, message, or idea.”127 In turn, the Eleventh Circuit reasoned that social refers to “a form of expression,”128 which “does not restrict any” subject matter or “type of speech.”129 Similarly, in NetChoice, LLC v. Bonta,130 the Ninth Circuit described the phrase “‘social media’ . . . as statutory shorthand” that does not implicate content distinctions.131 As of now, therefore, the law is unsettled on whether the “social” in “social media” constitutes a category of content under Reed and City of Austin.

2.  Content-Based Exemptions. — Courts have also held that social media age-verification laws are content based due to exemptions in coverage provisions. For example, the Northern District of Georgia stated that Georgia’s law’s “primary downfall . . . [was] its exemptions.”132 The law included exemptions for online platforms focused on “[e]mail,” “[n]ews, sports,” online gaming, “[c]loud storage,” and “advertising,” among others.133 The Arkansas,134 Texas,135 Louisiana,136 and Mississippi laws contained similar exemptions.137 One repeated exemption of note was for professional networking or career development services138 and is explored further in Part III. Content-based exemptions undermine states’ arguments that their laws should be subject to intermediate scrutiny because these exemptions are “based on the message a particular digital service provider conveys.”139

3.  Feature-Based Designations. — Florida’s law and California’s second law, the Protecting Our Kids from Social Media Addiction Act140 (POKSMA), are distinguishable from the above laws.141 Both define social media based on specific, technical features of platforms without targeting “social” content or listing extensive exemptions.

Florida’s law lists four criteria that determine whether a platform is covered: whether the platform (1) “[a]llows users to upload content,” (2) has “[t]en percent or more of . . . [its] daily active users . . . younger than 16 . . . spend on average 2 [or more] hours per day” on the platform, (3) “[e]mploys algorithms . . . to select content for users,” and (4) has any one of a number of “addictive features.”142 The law exempts platforms with “the exclusive function” of “e-mail or direct messaging” without public posting.143

The Northern District of Florida concluded that Florida’s law was not content based and applied intermediate scrutiny.144 In Uthmeier, the court found that under Turner the coverage provision was not content based as it was justified by a “special characteristic” of addictive social media platforms.145 The court also applied TikTok to recognize that the application of the law would not be altered if “the content of the speech on [the social media] platforms” changed.146 The Northern District of Georgia endorsed the same conclusion, distinguishing Georgia’s law from Florida’s on the basis that the latter did “not contain the same plethora of [content-based] exemptions” and instead regulated based on specific features.147 Upon review of an application for a stay of a preliminary injunction, the Eleventh Circuit followed the district court’s reasoning that the reliance on addictive features was content neutral.148

California also took a different approach to regulating minors’ social media use in POKSMA. Instead of restricting minors’ access to platforms, its law prohibits minors’ access to specific features of the platforms without parental consent.149 The law singles out “[a]ddictive feed[s]” — algorithm-based feeds that display content using individualized data150 — and displays of likes on public posts, and restricts the platforms from using these features for minors.151 The coverage provision extends to any online service with the specified features and does not make any additional definition of social media.152 The law also exempts platforms “limited to commercial transactions or to consumer reviews.”153

In Bonta, the Ninth Circuit held that POKSMA as a whole was not content based and applied intermediate scrutiny154 to uphold most challenged provisions of the law.155 The court dismissed the argument that the law’s focus on social media was content based, treating the term “‘social media’ platform as statutory shorthand” and reasoning that the coverage provision “applie[d] to websites” regardless of “whether they facilitate[d] social interaction.”156 The court also found that the exemptions for businesses were content neutral under City of Austin.157

III.  Potential Paths to a Content-Neutral Social Media Age-Verification Law

States can avoid strict scrutiny with content-neutral coverage provisions by focusing on the structural elements of the platforms they seek to regulate. Defining social media is a challenge for legislatures considering age-verification laws: A technical definition of social media is likely to encompass a wider set of platforms than those traditionally considered to be “social media.”158 Members of the Supreme Court have acknowledged this challenge.159 Legislators must be cognizant that overbroad legislation is constitutionally suspect and should therefore intelligibly scope social media age-verification laws without implicating “the vast majority of companies operating online.”160 Although selectively covering platforms may seem to require content-based distinctions, recent cases from the Supreme Court and lower courts suggest a constitutional path forward.

A.  The LinkedIn Problem: Defining Social Media Without Triggering Strict Scrutiny

LinkedIn presents the challenge of clearly defining social media platforms. While arguably not a traditional social media platform,161 it shares many of the features and structural elements of social media: Users create profiles, connect and communicate with one another, and look at algorithmically curated feeds of content. Indeed, the Supreme Court described LinkedIn as a paradigmatic example of social media in Packingham.162 Yet it seems a stretch to suggest that LinkedIn harms minors to an extent necessary to justify regulation by states.163 Accordingly, many states have adopted this position by excluding LinkedIn through exemptions for platforms that provide “[c]areer development opportunities.”164 A precocious fifteen-year-old who bypasses LinkedIn’s age restrictions with the goal of getting a head start building a professional network is likely not at risk of the harms age-verification laws seek to address.165

This “LinkedIn Problem” is something legislators must solve. If they wish to regulate platforms harmful to minors, they must do so without implicating relatively innocuous platforms, like LinkedIn, that nevertheless structurally resemble social media, or without using content-based distinctions. At first, the problem appears to be a catch-22: A law that covers platforms like LinkedIn will likely be overinclusive and struck down,166 but a law that distinguishes LinkedIn based on its professional content makes a content-based distinction, which requires strict scrutiny.167

Utah’s law ran into a variation of the LinkedIn Problem. In Net­Choice, LLC v. Reyes,168 the court held that Utah’s age-verification law was overinclusive for covering “social media platforms with[out] significant populations of minor users.”169 It noted that the law would cover Dreamwidth, a website used for sharing creative content in online communities.170 Dreamwidth is not the type of “traditional social media platform[]” states are trying to regulate in protecting minors171: The site has about 30,000 active users worldwide, fewer than 2,000 users below the age of eighteen, and fewer than 5,000 users in Utah.172 But the coverage provision of Utah’s law would require Dreamwidth to use age-verification technology given the law’s definition of a social media service.173

The Dreamwidth treatment illustrates the central importance of defining social media. It stretches credulity to think that an age-verification law is necessary to protect “the mental health [and] personal privacy” of the small number of minor Dreamwidth users in Utah.174 And the First Amendment is meant to protect private speakers, like Dreamwidth, from state regulation of their expression regardless of their size or power.175 The Reyes court’s suggestion that Dreamwidth is distinct from “traditional social media platforms [like] Facebook and X” acknowledges that there is something unique about the larger social media platforms that are the alleged drivers of harms to minors.176 Legislators must channel this acknowledgment into an articulable, content-neutral law that will merit intermediate scrutiny under the First Amendment.

B.  The TikTok Solution: Directly Naming Platforms

TikTok provides one potential solution to the LinkedIn Problem. Rather than crafting a complex definition of social media meant to apply only to platforms of concern or using content-based exemptions, states can directly name platforms in social media age-verification laws. For example, a state could specify that a law’s provisions apply to Facebook, Instagram, TikTok, Snapchat, and X. Congress did just this for the law challenged in TikTok.177 This strategy has its risks: Singling out a speaker based on the content of their speech is “presumptively unconstitutional” under the First Amendment.178 But TikTok stated that singling out a platform is not content based “when the differential treatment is ‘justified by some special characteristic of’ the” platform179 and reaffirmed Turner’s assertion that “regulation[s] that appl[y] to one medium (or a subset thereof) but not others” are not inherently content based.180

States can invoke TikTok and its reliance on Turner to argue they can regulate specific platforms due to their special characteristics. There’s surprisingly little scholarship or case law on what qualifies as a special characteristic, but TikTok and Turner relied on the scale of the regulated party and dangers associated with concentrated control over information or an entire medium of communication.181 The Court seems willing to identify special characteristics for large, sophisticated parties with both the capacity and incentive to control media in ways harmful to users.182

The addictive features of social media platforms, such as those identified in Florida’s law and POKSMA, likely satisfy the Court’s treatment of special characteristics. Social media platforms produce addiction-like behaviors, and almost half of all Americans describe themselves as “‘addicted’ to their smartphones.”183 Social media companies “knowingly use[]” addictive features,184 alongside their control of massive amounts of data,185 to facilitate addiction.186 While the harms of addiction-like behavior are cognizable for adults, they become even more acute with respect to minors,187 and courts seem willing to recognize the addictive effects of minors’ social media use.188 Scholars have suggested that regulations that facilitate “greater user choice” are “less restrictive” and more likely to be held constitutional.189 States can assert that the features causing addictive outcomes pose specific dangers to minors that justify their “differential treatment”190 without “undermining First Amendment interests.”191

States could alternatively designate specific platforms based on evidence of platforms’ causal connection to harms beyond addictive effects. For example, states could reference the Facebook Files — leaked documents illustrating Meta’s knowledge of the harmful effects of Instagram use by teenagers, including increases in suicidal thoughts, body image issues, and addictive behaviors192 — to conclude that Instagram merits special legislative treatment. Evidence of known harmful effects would support a claim that a “special characteristic”193 of a platform like Instagram justifies its singling out. States can alternatively make their own legislative determinations regarding which social media platforms employ features that are harmful to minors.

This approach comes with a set of potential concerns. First, targeting a small subset of traditionally recognized social media platforms may raise underinclusiveness concerns for failing to reach as much speech as needed to address the state’s interest.194 However, the TikTok Court clarified that a government does not need to “address all aspects of a problem in one fell swoop.”195 TikTok also does not give states a free pass to single out platforms with impunity: “[D]ifferential treatment [not] justified by some special characteristic” of a platform “is presumptively unconstitutional.”196

This solution to the LinkedIn Problem risks overextending TikTok, which explicitly noted “the inherent narrowness of [its] holding.”197 It is easy to envision the Court later clarifying that TikTok applied exclusively to its national security context and declining to extend its reasoning to other internet regulations. The TikTok Court also stated that “[a] law targeting any other speaker would by necessity entail a distinct inquiry and separate considerations.”198 So while TikTok may signal the Court’s willingness to allow direct regulation of platforms with special characteristics, states risk overextending the decision to social media age-verification laws.

C.  The Feature-Based Solution: Content-Neutral Descriptions of Social Media

The Florida and California laws described in section II.B.2 present an alternate solution. Both laws focus on the structural features of social media platforms and do not involve content-based designations. Florida defines social media platforms in terms of specified harmful features,199 and California bypasses the need to define social media platforms by regulating access to features rather than to entire platforms.200 States can follow these examples and write laws that merit only intermediate scrutiny by eliminating exemptions and targeting specific features, rather than employing a broader definition of “social interaction,” to define covered social media platforms.

The use of exemptions, while the simplest way to carve out platforms from broad coverage provisions, is incurably content based and will guarantee that social media age-verification laws are subject to strict scrutiny. Exemptions that elevate categories of content violate Reed’s admonishment against facially content-based regulations201 and the longstanding First Amendment norm that the government should not decide what content minors (or adults) engage with.202 While certain commercial exemptions may be acceptable under City of Austin,203 the use of exemptions risks a judicial presumption that states are engaged in content-based discrimination. States should abandon their attempts to distinguish the social media platforms they seek to cover with exemptions.

While courts’ conclusion that “social” is a content-based distinction is questionable,204 states can bypass this concern altogether by designating covered platforms based on their incorporation of specific harmful features such as algorithmic feeds based on individualized data, infinite scrolling, push notifications, or others. Indeed, there is significant and growing evidence that such features were deliberately designed to exploit human psychology.205 And courts have held that social media platforms may be liable for the harms these features bring to minors.206 States have good reasons to enact regulations based on these features given their exploitative nature, propensity to create harms, and deliberate inclusion in traditional social media platforms.207 Coverage provisions that define covered platforms by the presence of these features are content neutral because they do not “reflect[] a content preference.”208

Targeting harmful features rather than the platforms themselves has the added benefit of making laws more likely to satisfy heightened scrutiny. The Northern District of Florida in Uthmeier held that Florida’s law likely failed intermediate scrutiny, describing the age-verification requirement as “an extraordinarily blunt instrument” that “burden[ed] substantially more speech than [was] necessary.”209 But the Eleventh Circuit disagreed, emphasizing that the focus on “addictive features” that minors “are particularly susceptible to” made the law narrowly tailored.210 The Ninth Circuit’s similar analysis of narrow tailoring suggests that courts are more willing to uphold regulations of specific features rather than wholesale access to a platform. Such regulations “logically serve[] the end of protecting minors[]” without raising doubts about whether a state is using the regulation as “a backdoor means of controlling content.”211 This “relatively nuanced approach” to social media regulation212 is an avenue states should pursue as they try to find a content-neutral solution to the LinkedIn Problem.

Conclusion

The First Amendment imposes a high, but not insurmountable, hurdle for states to overcome in regulating minors’ social media use. By focusing on specific features that lead to harmful effects on minors, states can craft content-neutral laws that will merit only intermediate scrutiny. The solutions to the LinkedIn Problem proposed above — naming platforms directly under TikTok’s revival of the “special characteristics” standard or regulating specific harmful features without reference to content — are the two likeliest ways for states to have their laws upheld in court. Like California, states must be creative and flexible as they respond to a rapidly developing legal doctrine. If “[s]ocial media is a cancer on our society,”213 then seeking a constitutional cure is crucial even if current efforts “dwell only on the suffering of children.”214

Footnotes

  1. ^ See, e.g., Matt Barnum, Twelfth-Grade Math and Reading Scores in U.S. Hit New Low, Wall St. J. (Sep. 9, 2025, at 00:01 ET), https://www.wsj.com/us-news/education/test-scores-low-reading-math-naep-d87099b6 [https://perma.cc/62L8-MXUB].

    Return to citation ^

  2. ^ See, e.g., Joe Waters, Opinion, America’s Loneliness Epidemic Hits Kids Hardest, Newsweek (Mar. 4, 2024, at 10:37 ET), https://www.newsweek.com/americas-loneliness-epidemic-hits-kids-hardest-opinion-1874902 [https://perma.cc/G5ZF-9PCJ]; Ellen Barry, Americans’ Struggle with Mental Health, N.Y. Times (Aug. 8, 2024), https://www.nytimes.com/2024/08/08/briefing/mental-health-anxiety-depression.html [https://perma.cc/7B7X-EU3C].

    Return to citation ^

  3. ^ See, e.g., Press Release, Am. Psych. Ass’n, APA Report Calls on Social Media Companies to Take Responsibility to Protect Youth (Apr. 16, 2024), https://www.apa.org/news/press/releases/2024/04/social-media-companies-protect-youth [https://perma.cc/2KST-8XVF].

    Return to citation ^

  4. ^ 15 U.S.C. §§ 6501–6506.

    Return to citation ^

  5. ^ See Jonathan Haidt, The Anxious Generation 230 (2024); Georgia Wells & Jeff Horwitz, Facebook’s Effort to Attract Preteens Goes Beyond Instagram Kids, Documents Show, Wall St. J. (Sep. 28, 2021, at 13:24 ET), https://www.wsj.com/tech/facebook-instagram-kids-tweens-attract-11632849667 [https://perma.cc/CB2L-YTD5].

    Return to citation ^

  6. ^ Matt Brown & Matt O’Brien, Senate Strikes AI Regulatory Ban from GOP Bill After Uproar from the States, AP News (July 1, 2025, at 14:30 ET), https://apnews.com/article/congress-ai-provision-moratorium-states-20beeeb6967057be5fe64678f72f6ab0 [https://perma.cc/QV7W-MAQQ]. Senator Marsha Blackburn stated last summer that states “[a]re the ones that are protecting children in the virtual space.” Id.

    Return to citation ^

  7. ^ This Note uses “social media platforms” to describe websites and applications in accordance with the Supreme Court’s description in Moody v. NetChoice, LLC, 144 S. Ct. 2383 (2024). See id. at 2394–95.

    Return to citation ^

  8. ^ This Note uses “social media age-verification laws” to cover laws that either ban minors (or a subset of minors) from accessing platforms or restrict the features available to minors. These laws differ in meaningful ways, but the First Amendment issues raised by them are comparable, as evidenced by courts’ willingness to treat the cases together. See, e.g., NetChoice v. Carr, 789 F. Supp. 3d 1200, 1222 (N.D. Ga. 2025); NetChoice, LLC v. Fitch, 145 S. Ct. 2658, 2658 (2025) (Kavanaugh, J., concurring in the denial of the application to vacate stay).

    Return to citation ^

  9. ^ See, e.g., Noah Feldman, Opinion, The Age Verification Fallout on Free Speech Is Starting, Bloomberg (Aug. 26, 2025, at 07:00 ET), https://www.bloomberg.com/opinion/articles/2025-08-26/supreme-court-is-to-blame-for-bluesky-age-verification-fight-in-mississippi [https://perma.cc/W4UR-WGBZ].

    Return to citation ^

  10. ^ Reno v. ACLU, 521 U.S. 844, 882 (1997) (describing the burden a law requiring age verification for access to sexual material on the internet will have on the internet).

    Return to citation ^

  11. ^ See, e.g., NetChoice, LLC v. Fitch, 787 F. Supp. 3d 262, 283 (S.D. Miss. 2025). Mississippi’s law is in a unique posture: The Southern District of Mississippi issued a preliminary injunction of the law, id., but the Fifth Circuit issued an unexplained stay of the injunction, NetChoice, L.L.C. v. Fitch, No. 25-60348, 2025 WL 2078435, at *1 (5th Cir. July 17, 2025) (mem.) (per curiam). The Supreme Court denied NetChoice’s emergency application to vacate the stay. Fitch, 145 S. Ct. at 2658 (Kavanaugh, J., concurring in the denial of the application to vacate stay).

    Return to citation ^

  12. ^ Fitch, 145 S. Ct. at 2658 (Kavanaugh, J., concurring in the denial of the application to vacate stay) (stating that Mississippi’s law “likely violate[s]” the First Amendment but denying to vacate the stay due to “the balance of harms and equities”).

    Return to citation ^

  13. ^ Rebecca Aviel et al., From Gods to Google, 134 Yale L.J. 1269, 1300 (2025).

    Return to citation ^

  14. ^ Gerald Gunther, The Supreme Court, 1971 Term — Foreword: In Search of Evolving Doctrine on a Changing Court: A Model for a Newer Equal Protection, 86 Harv. L. Rev. 1, 8 (1972); accord Free Speech Coal., Inc. v. Paxton, 145 S. Ct. 2291, 2310 (2025). But see id. at 2320 (Kagan, J., dissenting) (stating that age-verification laws may be able to survive strict scrutiny).

    Return to citation ^

  15. ^ Paxton, 145 S. Ct. at 2317 (majority opinion) (quoting Turner Broad. Sys., Inc. v. FCC, 520 U.S. 180, 189 (1997)).

    Return to citation ^

  16. ^ See Sandy Dornsife, Eight States Enact Minor Social Media Bans Despite Court Fights, MultiState (Oct. 8, 2025), https://www.multistate.us/insider/2025/10/8/eight-states-enact-minor-social-media-bans-despite-court-fights [https://perma.cc/F8C9-WFR8] (describing repeated attempts by states to pass similar legislation).

    Return to citation ^

  17. ^ For example, Australia began enforcing its age-verification law in December 2025. Victoria Kim, Australia’s Social Media Ban for Children Takes Effect, N.Y. Times (Dec. 9, 2025), https://www.nytimes.com/2025/12/09/world/asia/australia-social-media-ban-under-16.html [https://perma.cc/KS3C-5G4Q].

    Return to citation ^

  18. ^ Contrast Haidt, supra note 5, at 148 & n.15 (stating that studies establish a causal link between social media and declines in minors’ mental health metrics), with Candice L. Odgers, The Great Rewiring, Unplugged, 628 Nature 29, 29 (2024) (reviewing Haidt, supra note 5) (stating that a causal link “is not supported by science”).

    Return to citation ^

  19. ^ See sources cited supra notes 1–2.

    Return to citation ^

  20. ^ Moody v. NetChoice, LLC, 144 S. Ct. 2383, 2403 (2024); see also id. at 2423 (Alito, J., concurring in the judgment) (“[R]esearch suggests that social media are having a devastating effect on many young people, leading to depression, isolation, bullying, and intense pressure to endorse the trend or cause of the day.”).

    Return to citation ^

  21. ^ See, e.g., Jeff Horwitz, Meta’s AI Rules Have Let Bots Hold “Sensual” Chats with Kids, Offer False Medical Info, Reuters (Aug. 14, 2025, at 06:00 GMT), https://www.reuters.com/investigates/special-report/meta-ai-chatbot-guidelines [https://perma.cc/4GRS-7UBM].

    Return to citation ^

  22. ^ This Note does not address whether states should enact age-verification laws to address the harms of minors’ social media use. Nor does it explore possible alternative legislative approaches. Rather, it adds to a growing literature on attempts to regulate minors’ social media access. See, e.g., Tanner Pool, Note, Honey, I Shrunk the Kids (Social Media Access): States’ Actions to Regulate Social Media Access for Minors Through Parental Consent and the First Amendment, 49 Okla. City U. L. Rev. 379, 385–86 (2025) (describing age-verification laws and proposing alternative solutions); Eric Goldman, The “Segregate-and-Suppress” Approach to Regulating Child Safety Online, 28 Stan. Tech. L. Rev. 173, 177–78 (2025) (criticizing states’ approaches to regulating social media access).

    Return to citation ^

  23. ^ 145 S. Ct. 57 (2025) (per curiam).

    Return to citation ^

  24. ^ U.S. Const. amend. I; see also Gitlow v. New York, 268 U.S. 652, 666 (1925) (incorporating the First Amendment against the states).

    Return to citation ^

  25. ^ See Sorrell v. IMS Health Inc., 564 U.S. 552, 565–66 (2011) (quoting United States v. Playboy Ent. Grp., Inc., 529 U.S. 803, 812 (2000)).

    Return to citation ^

  26. ^ Turner Broad. Sys., Inc. v. FCC, 512 U.S. 622, 642 (1994).

    Return to citation ^

  27. ^ R.A.V. v. City of St. Paul, 505 U.S. 377, 382 (1992).

    Return to citation ^

  28. ^ Regan v. Time, Inc., 468 U.S. 641, 648–49 (1984).

    Return to citation ^

  29. ^ Genevieve Lakier, Reed v. Town of Gilbert, Arizona, and the Rise of the Anticlassificatory First Amendment, 2016 Sup. Ct. Rev. 233, 238 (2017); see also W. Va. State Bd. of Educ. v. Barnette, 319 U.S. 624, 642 (1943) (“If there is any fixed star in our constitutional constellation, it is that no official . . . can prescribe what shall be orthodox in politics, nationalism, religion, or other matters of opinion . . . .”).

    Return to citation ^

  30. ^ Ward v. Rock Against Racism, 491 U.S. 781, 791 (1989). Ward held that “[t]he government’s purpose is the controlling consideration” when “determining content neutrality.” Id.

    Return to citation ^

  31. ^ 564 U.S. 552 (2011).

    Return to citation ^

  32. ^ Id. at 564–65.

    Return to citation ^

  33. ^ Id. at 564.

    Return to citation ^

  34. ^ 512 U.S. 622 (1994).

    Return to citation ^

  35. ^ Id. at 661–62.

    Return to citation ^

  36. ^ Id. at 658–59.

    Return to citation ^

  37. ^ Id. at 661.

    Return to citation ^

  38. ^ See id. at 676–78 (O’Connor, J., concurring in part and dissenting in part).

    Return to citation ^

  39. ^ Evelyn Douek & Genevieve Lakier, The Supreme Court, 2023 Term — Comment: Lochner.com?, 138 Harv. L. Rev. 100, 121 (2024).

    Return to citation ^

  40. ^ Id. at 139.

    Return to citation ^

  41. ^ 576 U.S. 155 (2015).

    Return to citation ^

  42. ^ Id. at 163; accord id. at 164.

    Return to citation ^

  43. ^ Id. at 170 (quoting Turner, 512 U.S. at 658).

    Return to citation ^

  44. ^ See Lakier, supra note 29, at 254.

    Return to citation ^

  45. ^ 142 S. Ct. 1464 (2022).

    Return to citation ^

  46. ^ Id. at 1474 (quoting Reed, 576 U.S. at 171); accord Aziz Z. Huq, The Trouble with Classifications, 100 Notre Dame L. Rev. 1, 21 (2024).

    Return to citation ^

  47. ^ City of Austin, 142 S. Ct. at 1472.

    Return to citation ^

  48. ^ Huq, supra note 46, at 39.

    Return to citation ^

  49. ^ See, e.g., NetChoice, LLC v. Yost, 778 F. Supp. 3d 923, 952 (S.D. Ohio 2025).

    Return to citation ^

  50. ^ See City of Austin, 142 S. Ct. at 1474.

    Return to citation ^

  51. ^ E.g., Douek & Lakier, supra note 39, at 100.

    Return to citation ^

  52. ^ Cf. Stephanie D’Abruzzo & Rick Lyon, The Internet Is for Porn, on Avenue Q: The Musical: Original Broadway Cast Recording (CD, BMG Music Oct. 6, 2003) (making light of the ease of accessing pornography on the internet).

    Return to citation ^

  53. ^ See Communications Decency Act of 1996 (CDA), 47 U.S.C. § 223(d) (criminalizing the transmission of content that would be obscene for minors), invalidated by, Reno v. ACLU, 521 U.S. 844 (1997); Child Online Protection Act (COPA), 47 U.S.C. § 231(a)(1), (c)(1)(B) (criminalizing the online posting of “material that is harmful to minors,” id. § 231(a)(1), and including an age-verification defense), invalidated by, ACLU v. Mukasey, 534 F.3d 181 (3d Cir. 2008), cert. denied, 555 U.S. 1137 (2009).

    Return to citation ^

  54. ^ 521 U.S. 844 (1997).

    Return to citation ^

  55. ^ 542 U.S. 656 (2004).

    Return to citation ^

  56. ^ Reno, 521 U.S. at 868 (holding the CDA was content based because its “purpose . . . [wa]s to protect children from . . . ‘indecent’ . . . speech” (quoting 47 U.S.C. § 223(a)(1)(A) (1994))); Ashcroft II, 542 U.S. at 665 (accepting the district court’s conclusion that COPA was content based). The Court held that both laws violated the First Amendment. Reno, 521 U.S. at 849; Ashcroft II, 542 U.S. at 670. Before Free Speech Coalition, Inc. v. Paxton, 145 S. Ct. 2291 (2025), Ashcroft II was the last time the Court addressed the constitutionality of online age-verification requirements. See id. at 2306 (“Ashcroft II has been [the Court’s] last word on the government’s power to protect children from sexually explicit content online.”).

    Return to citation ^

  57. ^ 582 U.S. 98 (2017).

    Return to citation ^

  58. ^ Id. at 109.

    Return to citation ^

  59. ^ Id. at 101.

    Return to citation ^

  60. ^ Id. at 107.

    Return to citation ^

  61. ^ Id. at 104, 107.

    Return to citation ^

  62. ^ See, e.g., id. at 110 (Alito, J., concurring in the judgment); Mary Anne Franks, Beyond the Public Square: Imagining Digital Democracy, 131 Yale L.J.F. 427, 429 (2021); see also Jack M. Balkin, Moody v. NetChoice: The Supreme Court Meets the Free Speech Triangle, 2024 Sup. Ct. Rev. 127, 136 (2025) (stating that “First Amendment doctrine is ill-suited to deal with” internet speech issues).

    Return to citation ^

  63. ^ Franks, supra note 62, at 427.

    Return to citation ^

  64. ^ Packingham, 582 U.S. at 107–08.

    Return to citation ^

  65. ^ 144 S. Ct. 2383 (2024).

    Return to citation ^

  66. ^ Id. at 2405.

    Return to citation ^

  67. ^ See Douek & Lakier, supra note 39, at 134 & n.30.

    Return to citation ^

  68. ^ Moody, 144 S. Ct. at 2399; Douek & Lakier, supra note 39, at 134.

    Return to citation ^

  69. ^ Cf. Douek & Lakier, supra note 39, at 138–39 (suggesting that by declining to answer this question, the Court indicated it may be willing to find regulations that implicate internet content to be content neutral if justified for different purposes); see also Balkin, supra note 62, at 141 (stating that Florida’s law challenged in Moody was content based and the Court could have held the law unconstitutional on this basis rather than remanding).

    Return to citation ^

  70. ^ 145 S. Ct. 2291 (2025).

    Return to citation ^

  71. ^ TikTok Inc. v. Garland, 145 S. Ct. 57, 70 (2025) (per curiam); Paxton, 145 S. Ct. at 2319.

    Return to citation ^

  72. ^ TikTok, 145 S. Ct. at 68–69.

    Return to citation ^

  73. ^ Paxton, 145 S. Ct. at 2314; accord id. at 2319.

    Return to citation ^

  74. ^ Douek & Lakier, supra note 39, at 104 n.30. Scholars compare modern First Amendment jurisprudence to due process holdings during the Lochner era of the early nineteenth century when the Fourteenth Amendment was used to invalidate a myriad of regulations. See, e.g., id. at 104 & n.30; Jeremy K. Kessler & David E. Pozen, Introduction, The Search for an Egalitarian First Amendment, 118 Colum. L. Rev. 1953, 1962 (2018).

    Return to citation ^

  75. ^ TikTok, 145 S. Ct. at 69.

    Return to citation ^

  76. ^ Id. at 67 (citation omitted).

    Return to citation ^

  77. ^ Id. at 69.

    Return to citation ^

  78. ^ Id. at 68 (quoting Turner Broad. Sys., Inc. v. FCC, 512 U.S. 622, 660–61 (1994)).

    Return to citation ^

  79. ^ Id. (quoting Turner, 512 U.S. at 658). The Court did not address concerns that TikTok would promote views endorsed by the Chinese government or evidence that legislators supported the law to limit the influence of pro-Palestinian content. Brief for Petitioners at 16, TikTok, 145 S. Ct. 57 (No. 24-656).

    Return to citation ^

  80. ^ TikTok, 145 S. Ct. at 67 (quoting Turner, 512 U.S. at 644).

    Return to citation ^

  81. ^ See The Supreme Court, 2024 Term — Leading Case: TikTok Inc. v. Garland, 139 Harv. L. Rev. 280, 286 (2025).

    Return to citation ^

  82. ^ TikTok, 145 S. Ct. at 68 (quoting Turner, 512 U.S. at 660–61). TikTok cited Reed only four times while citing Turner and its follow-up case Turner Broadcasting System, Inc. v. FCC, 520 U.S. 180 (1997), fourteen times.

    Return to citation ^

  83. ^ Free Speech Coal., Inc. v. Paxton, 145 S. Ct. 2291, 2319 (2025).

    Return to citation ^

  84. ^ Bd. of Educ. v. Pico, 457 U.S. 853, 867 (1982) (opinion of Brennan, J.) (quoting Stanley v. Georgia, 394 U.S. 557, 564 (1969)).

    Return to citation ^

  85. ^ Erznoznik v. City of Jacksonville, 422 U.S. 205, 212 (1975).

    Return to citation ^

  86. ^ 390 U.S. 629 (1968).

    Return to citation ^

  87. ^ Id. at 637. The Court later read Ginsberg as upholding age verification. See Paxton, 145 S. Ct. at 2307–08.

    Return to citation ^

  88. ^ Brown v. Ent. Merchs. Ass’n, 564 U.S. 786, 792 (2011); see also id. at 799 (holding that singling out “violent video games” was a content-based restriction).

    Return to citation ^

  89. ^ Paxton, 145 S. Ct. at 2319.

    Return to citation ^

  90. ^ Texas’s law covers websites where “more than one-third of [the content] is sexual material harmful to minors.” Tex. Civ. Prac. & Rem. Code Ann. § 129B.002(a) (West 2025). The law defines what content is “[s]exual material harmful to minors.” Id. § 129B.001(6).

    Return to citation ^

  91. ^ Paxton, 145 S. Ct. at 2306 (quoting Boy Scouts of Am. v. Dale, 530 U.S. 640, 659 (2000)). In what was arguably dicta, the Paxton Court stated that “adults have no First Amendment right to avoid age verification.” Id. at 2309. This line will likely be fought over in litigation regarding social media age-verification cases. See, e.g., Reply in Support of Emergency Application of NetChoice for Immediate Temporary Administrative Relief and Vacatur of Fifth Circuit’s Stay of Preliminary Injunction at 11, NetChoice, LLC v. Fitch, 145 S. Ct. 2658 (2025) (No. 25A97).

    Return to citation ^

  92. ^ Paxton, 145 S. Ct. at 2306 (citing Ginsberg, 390 U.S. at 641).

    Return to citation ^

  93. ^ See Douek & Lakier, supra note 39, at 108.

    Return to citation ^

  94. ^ See US State Age Assurance Laws for Social Media, Age Verification Providers Ass’n (Oct. 2025), https://avpassociation.com/us-state-age-assurance-laws-for-social-media [https://perma.cc/BME6-8RP6].

    Return to citation ^

  95. ^ See, e.g., NetChoice, LLC v. Yost, 778 F. Supp. 3d 923, 959 (S.D. Ohio 2025); NetChoice v. Carr, 789 F. Supp. 3d 1200, 1210 (N.D. Ga. 2025). The Ninth Circuit affirmed and partially reversed preliminary injunctions of California’s two laws. See infra note 141. The Eleventh and Fifth Circuits also issued stays of the preliminary injunctions against Florida’s and Mississippi’s laws, respectively. See Comput. & Commc’ns Indus. Ass’n v. Uthmeier, No. 25-11881, 2025 WL 3458571, at *1 (11th Cir. Nov. 25, 2025); supra note 11.

    Return to citation ^

  96. ^ This Part focuses on laws addressed by courts.

    Return to citation ^

  97. ^ Utah Code Ann. § 13-71-101(14)(a) (West 2025); accord id. § 13-71-101(14)(a)(i)–(v).

    Return to citation ^

  98. ^ Ohio Rev. Code Ann. § 1349.09(B) (West 2025); see also NetChoice, LLC v. Bonta, 770 F. Supp. 3d 1164, 1179 (N.D. Cal. 2025) (quoting NetChoice, LLC v. Bonta, 113 F.4th 1101, 1109 (9th Cir. 2024)).

    Return to citation ^

  99. ^ Carr, 789 F. Supp. 3d at 1219 (citing Ga. Code Ann. § 39-6-1(6) (West 2025)).

    Return to citation ^

  100. ^ Id. at 1220.

    Return to citation ^

  101. ^ Miss. Code Ann. § 45-38-5(2)(c)(i) (West 2025); accord Tex. Bus. & Com. Code Ann. § 509.002(b)(1)(A) (West 2025).

    Return to citation ^

  102. ^ Cal. Health & Safety Code § 27000.5(b)(2)(A) (West 2025).

    Return to citation ^

  103. ^ E.g., Miss. Code Ann. § 45-38-5(2)(c)(i).

    Return to citation ^

  104. ^ For the difference between age verification and age assurance, see Sarah Forland, Nat Meysenburg & Erika Solis, New Am., Age Verification: The Complicated Effort to Protect Youth Online 10 (2024), https://d1y8sb8igg2f8e.cloudfront.net/documents/Age_Verification_The_Complicated_Effort_to_Protect_Youth_Online_2024-04-22_165_bS2AcQ5.pdf [https://perma.cc/2CB6-QK9L].

    Return to citation ^

  105. ^ Ark. Code Ann. § 4-88-1402(c)(1) (West 2025). Other states have similar provisions. E.g., Miss. Code Ann. § 45-38-7(1).

    Return to citation ^

  106. ^ Utah Code Ann. § 13-71-101(2) (West 2025).

    Return to citation ^

  107. ^ See, e.g., Ohio Rev. Code Ann. § 1349.09(B)(1) (West 2025).

    Return to citation ^

  108. ^ See, e.g., NetChoice, LLC v. Yost, 716 F. Supp. 3d 539, 558 n.9 (S.D. Ohio 2024).

    Return to citation ^

  109. ^ See, e.g., Fla. Stat. Ann. § 501.1736(2)(a) (West 2025) (prohibiting minors under fourteen from becoming account holders).

    Return to citation ^

  110. ^ See, e.g., Ohio Rev. Code Ann. § 1349.09(E).

    Return to citation ^

  111. ^ See, e.g., Cal. Health & Safety Code §§ 27001(a), 27002 (West 2025).

    Return to citation ^

  112. ^ See, e.g., Utah Code Ann. § 13-71-301(3)(a)(i) (West 2025).

    Return to citation ^

  113. ^ See, e.g., Tex. Bus. & Com. Code Ann. § 509.152(b) (West 2025).

    Return to citation ^

  114. ^ § 1349.09(A)(2).

    Return to citation ^

  115. ^ Id. § 1349.09(B)(1).

    Return to citation ^

  116. ^ Health & Safety § 27001(a)(2).

    Return to citation ^

  117. ^ Id. § 27002(b).

    Return to citation ^

  118. ^ Bus. & Com. § 509.051(d)(2)(B).

    Return to citation ^

  119. ^ Id. §§ 509.001(7), 509.101.

    Return to citation ^

  120. ^ See, e.g., Defendants’ Memorandum in Opposition to Plaintiff’s Motion for Preliminary Injunction at 41–42, NetChoice, LLC v. Reyes, 748 F. Supp. 3d 1105 (D. Utah 2024) (No. 23-cv-00911).

    Return to citation ^

  121. ^ Cases addressing Florida’s law and California’s second law are the two exceptions. See infra notes 126–31 and accompanying text; see also infra section II.B.3, pp. 954–56.

    Return to citation ^

  122. ^ NetChoice, LLC v. Yost, 716 F. Supp. 3d 539, 557 (S.D. Ohio 2024).

    Return to citation ^

  123. ^ Id.

    Return to citation ^

  124. ^ Reyes, 748 F. Supp. 3d at 1122 (quoting NetChoice, LLC v. Fitch, 738 F. Supp. 3d 753, 770 (S.D. Miss. 2024)).

    Return to citation ^

  125. ^ E.g., Comput. & Commc’ns Indus. Ass’n v. Paxton, 747 F. Supp. 3d 1011, 1032 (W.D. Tex. 2024).

    Return to citation ^

  126. ^ No. 24-cv-00438, 2025 WL 1570007 (N.D. Fla. June 3, 2025).

    Return to citation ^

  127. ^ Id. at *14.

    Return to citation ^

  128. ^ Comput. & Commc’ns Indus. Ass’n v. Uthmeier, No. 25-11881, 2025 WL 3458571, at *4 (11th Cir. Nov. 25, 2025).

    Return to citation ^

  129. ^ Id. at *5.

    Return to citation ^

  130. ^ 152 F.4th 1002 (9th Cir. 2025).

    Return to citation ^

  131. ^ Id. at 1016; see also NetChoice v. Carr, 789 F. Supp. 3d 1200, 1219–20 (N.D. Ga. 2025) (stating that “[a]t first glance . . . [Georgia’s] definition [of ‘social media platform’] does not appear to reference the content of the[] platforms,” id. at 1219, before proceeding to conclude that the law’s exemptions were content based).

    Return to citation ^

  132. ^ Carr, 789 F. Supp. 3d at 1219.

    Return to citation ^

  133. ^ Id. at 1219–20 (quoting Ga. Code Ann. § 39-6-1(6) (West 2025)).

    Return to citation ^

  134. ^ See Ark. Code Ann. § 4-88-1401(11)(B) (West 2025).

    Return to citation ^

  135. ^ See Tex. Bus. & Com. Code Ann. § 509.002(b)(10)(A) (West 2025).

    Return to citation ^

  136. ^ See La. Stat. Ann. § 51:1751(12)(b) (2025).

    Return to citation ^

  137. ^ See Miss. Code Ann. § 45-38-5(2) (West 2025).

    Return to citation ^

  138. ^ E.g., Ga. Code Ann. § 39-6-1(6)(V) (excluding platforms with the primary function of “[c]areer development opportunities”); Comput. & Commc’ns Indus. Ass’n v. Paxton, 747 F. Supp. 3d 1011, 1032 (W.D. Tex. 2024) (stating that “professional interactions[] are not covered” by Texas’s law).

    Return to citation ^

  139. ^ NetChoice, LLC v. Fitch, 787 F. Supp. 3d 262, 276 (S.D. Miss. 2025); cf. Williams-Yulee v. Fla. Bar, 575 U.S. 433, 449 (2015) (expressing skepticism of speech-related laws that are “riddled with exceptions”). The law challenged in TikTok included exemptions for websites “whose primary purpose is to allow users to post product reviews, business reviews, or travel information and reviews.” TikTok Inc. v. Garland, 145 S. Ct. 57, 64 (2025) (per curiam) (quoting Protecting Americans from Foreign Adversary Controlled Applications Act (TikTok Control Act), Pub. L. No. 118-50, § 2(g)(2)(B), 138 Stat. 955, 958 (2024)). The Supreme Court did not address the exemptions because the law was challenged only as applied to TikTok. Id. at 68. However, TikTok’s arguments mirrored those made in challenges to age-verification laws. See, e.g., Brief for Petitioners, supra note 79, at 44 (“Congress’s adoption of content-based exceptions strongly suggests that content rather than data is what motivated it.”).

    Return to citation ^

  140. ^ Cal. Health & Safety Code § 27000 (West 2025).

    Return to citation ^

  141. ^ California first passed the California Age-Appropriate Design Code Act (CAADCA), Cal. Civ. Code §§ 1798.99.28–1798.99.40 (West 2025), in 2022. Jesús Alvarado & Dean Jackson, The California Age Appropriate Design Code Act May Be the Most Important Piece of Tech Legislation You’ve Never Heard of, Tech Pol’y Press (July 9, 2024), https://www.techpolicy.press/the-california-age-appropriate-design-code-act-may-be-the-most-important-piece-of-tech-legislation-youve-never-heard-of [https://perma.cc/5UQS-VM28]. The Act was enjoined first in NetChoice, LLC v. Bonta, 692 F. Supp. 3d 924, 966 (N.D. Cal. 2023), aff’d in part, vacated in part, 113 F.4th 1101 (9th Cir. 2024), and then enjoined again on remand in NetChoice, LLC v. Bonta, 770 F. Supp. 3d 1164, 1215 (N.D. Cal. 2025). California then passed POKSMA in 2024. Lindsey Tonsager et al., California Passes Law to Protect Minors from “Addictive Feeds,” Covington: Inside Priv. (Oct. 10, 2024), https://www.insideprivacy.com/uncategorized/california-passes-law-to-protect-minors-from-addictive-feeds [https://perma.cc/XY2L-T4X4]. The Northern District of California issued a partial preliminary injunction of the law, NetChoice v. Bonta, 761 F. Supp. 3d 1202, 1210 (N.D. Cal. 2024), which the Ninth Circuit largely affirmed, NetChoice, LLC v. Bonta, 152 F.4th 1002, 1009 (9th Cir. 2025).

    Return to citation ^

  142. ^ Fla. Stat. Ann. § 501.1736(1)(e) (West 2025). The listed addictive features are “[i]nfinite scrolling,” “[p]ush notifications,” “[d]isplays [of] personal interactive metrics” (such as like or repost counts), “[a]uto-play video[s],” and “[l]ive-streaming.” Id.

    Return to citation ^

  143. ^ Id.

    Return to citation ^

  144. ^ Comput. & Commc’ns Indus. Ass’n v. Uthmeier, No. 24-cv-00438, 2025 WL 1570007, at *13 (N.D. Fla. June 3, 2025).

    Return to citation ^

  145. ^ Id. (quoting Turner Broad. Sys., Inc. v. FCC, 512 U.S. 622, 660–61 (1994)).

    Return to citation ^

  146. ^ Id. (“[P]latforms . . . ‘cannot avoid or mitigate’ the effects of the law . . . by changing their content moderation policies . . . .” (quoting TikTok Inc. v. Garland, 145 S. Ct. 57, 67 (2025) (per curiam))). The court concluded that Florida’s law did not meet intermediate scrutiny because it burdened more speech than necessary. Id. at *15.

    Return to citation ^

  147. ^ NetChoice v. Carr, 789 F. Supp. 3d 1200, 1222 (N.D. Ga. 2025).

    Return to citation ^

  148. ^ Comput. & Commc’ns Indus. Ass’n v. Uthmeier, No. 25-11881, 2025 WL 3458571, at *4 (11th Cir. Nov. 25, 2025).

    Return to citation ^

  149. ^ NetChoice, LLC v. Bonta, 152 F.4th 1002, 1010 (9th Cir. 2025).

    Return to citation ^

  150. ^ Cal. Health & Safety Code § 27000.5(a) (West 2025).

    Return to citation ^

  151. ^ Id. § 27002(b). POKSMA includes an age-verification requirement that goes into effect in 2027. Id. §§ 27001, 27006(b). The Ninth Circuit held that the challenge to the age-verification requirement was not ripe. Bonta, 152 F.4th at 1019.

    Return to citation ^

  152. ^ § 27000.5(b)(1).

    Return to citation ^

  153. ^ Id. § 27000.5(b)(2)(A).

    Return to citation ^

  154. ^ See Bonta, 152 F.4th at 1015.

    Return to citation ^

  155. ^ Id. at 1009.

    Return to citation ^

  156. ^ Id. at 1016.

    Return to citation ^

  157. ^ Id. at 1015–16. The Ninth Circuit held that the provision barring the display of “like counts” to minors was content based, id. at 1016, and failed strict scrutiny but was severable from the Act, id. at 1024.

    Return to citation ^

  158. ^ See Thomas Aichner et al., Twenty-Five Years of Social Media: A Review of Social Media Applications and Definitions from 1994 to 2019, 24 Cyberpsychology Behav. & Soc. Network­ing 215, 215 (2021) (“[Social media] is generally used as an umbrella term that describes a variety of online platforms . . . .”); see also NetChoice, LLC v. Griffin, 23-CV-5105, 2025 WL 978607, at *2 (W.D. Ark. Mar. 31, 2025) (stating that Arkansas’s definition of social media doesn’t “follow[] common parlance”).

    Return to citation ^

  159. ^ See, e.g., Packingham v. North Carolina, 582 U.S. 98, 118 n.16 (2017) (Alito, J., concurring in the judgment) (“[I]t is not easy to provide a precise definition of a ‘social media’ site . . . .”).

    Return to citation ^

  160. ^ NetChoice, LLC v. Bonta, 770 F. Supp. 3d 1164, 1179 (N.D. Cal. 2025).

    Return to citation ^

  161. ^ LinkedIn describes itself as a “professional network.” About LinkedIn, LinkedIn, https://about.linkedin.com [https://perma.cc/CG8X-5H9A]. But see Samantha Jevons, Why LinkedIn Is Evol­ving into a Personal Social Media Platform, LinkedIn (Oct. 2, 2024), https://www.linkedin.com/pulse/why-linkedin-evolving-personal-social-media-platform-samantha-jevons-uldxf [https://perma.cc/F7AM-47JR].

    Return to citation ^

  162. ^ Packingham, 582 U.S. at 106–07 (describing Facebook, LinkedIn, and Twitter as websites that are “commonly understood,” id. at 106 (quoting Brief of Respondent at 54, Packingham, 582 U.S. 98 (No. 15-1194), 2017 WL 345120, at *54), to be social media).

    Return to citation ^

  163. ^ In fact, LinkedIn’s age requirement of sixteen is higher than that of most platforms. FAQs for Age-Based Protections on LinkedIn, LinkedIn (2024), https://www.linkedin.com/help/linkedin/answer/a6854067 [https://perma.cc/6RK5-FGDG]. Most platforms require users to be thirteen. See, e.g., About Instagram Teen Privacy and Safety Settings, Instagram, https://help.instagram.com/3237561506542117 [https://perma.cc/7RC5-N6J2].

    Return to citation ^

  164. ^ Ga. Code Ann. § 39-6-1(6)(V) (West 2025); see also, e.g., Miss. Code Ann. § 45-38-5(2)(d) (West 2025).

    Return to citation ^

  165. ^ Cf. Anya Kamenetz, Teens Love LinkedIn, The Cut (Oct. 6, 2023), https://www.thecut.com/article/why-teenagers-love-linkedin.html [https://perma.cc/7HUH-FP8K] (illustrating how minors use LinkedIn for professional networking and for traditional social media uses).

    Return to citation ^

  166. ^ See, e.g., NetChoice, LLC v. Fitch, 787 F. Supp. 3d 262, 278–79 (S.D. Miss. 2025).

    Return to citation ^

  167. ^ See supra p. 953.

    Return to citation ^

  168. ^ 748 F. Supp. 3d 1105 (D. Utah 2024).

    Return to citation ^

  169. ^ Id. at 1129.

    Return to citation ^

  170. ^ Id.; Dreamwidth, https://www.dreamwidth.org [https://perma.cc/NQS3-AJDJ]. The court described Dreamwidth as a “social networking” site. Reyes, 748 F. Supp. 3d at 1129 (quoting Declaration of Denise Paolucci in Support of Plaintiff’s Motion for Preliminary Injunction ¶ 2, Reyes, 748 F. Supp. 3d 1105 (No. 23-cv-00911)). Dreamwidth does not explicitly describe itself as a social media platform. See What Is Dreamwidth Studios?, Dreamwidth (Aug. 30, 2015), https://www.dreamwidth.org/support/faqbrowse?faqid=1 [https://perma.cc/V5MB-HEX6] (describing Dreamwidth as “part social network, part blogging system, and part content management system”); see also NetChoice, LLC v. Yost, 778 F. Supp. 3d 923, 947 (S.D. Ohio 2025) (listing Dreamwidth as an example of a covered platform under Ohio’s law).

    Return to citation ^

  171. ^ Reyes, 748 F. Supp. 3d at 1129.

    Return to citation ^

  172. ^ See Dreamwidth Statistics, Dreamwidth, https://www.dreamwidth.org/stats [https://perma.cc/9F8F-Y5Z5]. The population of Utah is approximately 3.48 million, and the population of minors in Utah is approximately 916,000. See 2025 Demographics, Healthy Salt Lake (May 2025), https://www.healthysaltlake.org/demographicdata [https://perma.cc/B3D8-KGGF]. Dreamwidth also does not promote its platform to minors. Reyes, 748 F. Supp. 3d at 1129.

    Return to citation ^

  173. ^ Reyes, 748 F. Supp. 3d at 1129; see also supra note 106 and accompanying text.

    Return to citation ^

  174. ^ Reyes, 748 F. Supp. 3d at 1124; accord id. at 1125.

    Return to citation ^

  175. ^ See Catharine A. MacKinnon, Weaponizing the First Amendment: An Equality Reading, 106 Va. L. Rev. 1223, 1229–30 (2020).

    Return to citation ^

  176. ^ Reyes, 748 F. Supp. 3d at 1129.

    Return to citation ^

  177. ^ See generally TikTok Control Act, Pub. L. No. 118-50, 138 Stat. 955 (2024).

    Return to citation ^

  178. ^ Reed v. Town of Gilbert, 576 U.S. 155, 163 (2015).

    Return to citation ^

  179. ^ TikTok Inc. v. Garland, 145 S. Ct. 57, 68 (2025) (per curiam) (quoting Turner Broad. Sys., Inc. v. FCC, 512 U.S. 622, 660–61 (1994)).

    Return to citation ^

  180. ^ Turner, 512 U.S. at 660. The D.C. Circuit explicitly quoted this text, TikTok Inc. v. Garland, 122 F.4th 930, 951 (D.C. Cir. 2024) (quoting Turner, 512 U.S. at 660), and quoted BellSouth Corp. v. FCC, 144 F.3d 58 (D.C. Cir. 1998), TikTok, 122 F.4th at 951 (stating that a statute does not automatically “warrant[] strict First Amendment review because it targets named corporations” (quoting BellSouth, 144 F.3d at 68)).

    Return to citation ^

  181. ^ See TikTok, 145 S. Ct. at 68; Turner, 512 U.S. at 661.

    Return to citation ^

  182. ^ See Balkin, supra note 62, at 151.

    Return to citation ^

  183. ^ Matthew B. Lawrence, Addiction and Liberty, 108 Corn. L. Rev. 259, 290 (2023) (quoting Trevor Wheelwright, 2022 Cell Phone Usage Statistics: How Obsessed Are We?, Reviews.org (Jan. 24, 2022), https://www.reviews.org/mobile/2022-cell-phone-addiction [https://perma.cc/FZS7-57YY]).

    Return to citation ^

  184. ^ Id. at 292.

    Return to citation ^

  185. ^ Balkin, supra note 62, at 148.

    Return to citation ^

  186. ^ Lawrence, supra note 183, at 291–92.

    Return to citation ^

  187. ^ See Haidt, supra note 5, at 5.

    Return to citation ^

  188. ^ See, e.g., Comput. & Commc’ns Indus. Ass’n v. Uthmeier, No. 24-cv-00438, 2025 WL 1570007, at *15 (N.D. Fla. June 3, 2025).

    Return to citation ^

  189. ^ Balkin, supra note 62, at 149.

    Return to citation ^

  190. ^ Turner Broad. Sys., Inc. v. FCC, 512 U.S. 622, 660 (1994).

    Return to citation ^

  191. ^ Id. at 661.

    Return to citation ^

  192. ^ Georgia Wells, Jeff Horwitz & Deepa Seetharaman, Facebook Knows Instagram Is Toxic for Teen Girls, Company Documents Show, Wall St. J. (Sep. 14, 2021, at 07:59 ET), https://www.wsj.com/tech/personal-tech/facebook-knows-instagram-is-toxic-for-teen-girls-company-documents-show-11631620739 [https://perma.cc/7YJ9-JFTY].

    Return to citation ^

  193. ^ TikTok Inc. v. Garland, 145 S. Ct. 57, 68 (2025) (per curiam) (quoting Turner, 512 U.S. at 660–61).

    Return to citation ^

  194. ^ Clay Calvert, Underinclusivity and the First Amendment: The Legislative Right to Nibble at Problems After Williams-Yulee, 48 Ariz. St. L.J. 525, 528 (2016).

    Return to citation ^

  195. ^ TikTok, 145 S. Ct. at 70 (quoting Williams-Yulee v. Fla. Bar, 575 U.S. 433, 449 (2015)).

    Return to citation ^

  196. ^ Minneapolis Star & Trib. Co. v. Minn. Comm’r of Revenue, 460 U.S. 575, 585 (1983); see also id. at 590 (stating that Minnesota “offered no adequate justification for the special treatment of” the singled-out medium); TikTok, 145 S. Ct. at 68.

    Return to citation ^

  197. ^ TikTok, 145 S. Ct. at 68; accord id. at 62–63.

    Return to citation ^

  198. ^ Id. at 69.

    Return to citation ^

  199. ^ See supra p. 955.

    Return to citation ^

  200. ^ See supra p. 956.

    Return to citation ^

  201. ^ See supra notes 41–44 and accompanying text.

    Return to citation ^

  202. ^ One exception is the obscene-for-minors category of content that doesn’t implicate minors’ free speech right. Free Speech Coal., Inc. v. Paxton, 145 S. Ct. 2291, 2306 (2025).

    Return to citation ^

  203. ^ NetChoice, LLC v. Bonta, 152 F.4th 1002, 1015–16 (9th Cir. 2025) (discussing City of Austin v. Reagan Nat’l Advert. of Aus., LLC, 142 S. Ct. 1464, 1473–74 (2022)).

    Return to citation ^

  204. ^ See supra p. 953.

    Return to citation ^

  205. ^ See Haidt, supra note 5, at 60–61.

    Return to citation ^

  206. ^ See Anderson v. TikTok, Inc., 116 F.4th 180, 184 (3d Cir. 2024) (holding that 47 U.S.C. § 230 does not immunize TikTok from tort claims resulting from its own algorithmic feeds); see also Doe v. Snap, Inc., 144 S. Ct. 2493, 2494 (2024) (Thomas, J., dissenting from the denial of certiorari) (questioning whether platforms can simultaneously claim First Amendment rights for their content and disclaim liability for resulting harms).

    Return to citation ^

  207. ^ See Haidt, supra note 5, at 229.

    Return to citation ^

  208. ^ Reed v. Town of Gilbert, 576 U.S. 155, 170 (2015) (quoting Turner Broad. Sys., Inc. v. FCC, 512 U.S. 622, 658 (1994)).

    Return to citation ^

  209. ^ Comput. & Commc’ns Indus. Ass’n v. Uthmeier, No. 24-cv-00438, 2025 WL 1570007, at *15 (N.D. Fla. June 3, 2025) (quoting Packingham v. North Carolina, 582 U.S. 98, 106 (2017)).

    Return to citation ^

  210. ^ Comput. & Commc’ns Indus. Ass’n v. Uthmeier, No. 25-11881, 2025 WL 3458571, at *6 (11th Cir. Nov. 25, 2025).

    Return to citation ^

  211. ^ NetChoice, LLC v. Bonta, 152 F.4th 1002, 1017 (9th Cir. 2025).

    Return to citation ^

  212. ^ Id. at 1018.

    Return to citation ^

  213. ^ Nathan Taylor Pemberton, Opinion, Charlie Kirk’s Killing and Our Poisonous Internet, N.Y. Times (Sep. 14, 2025) (quoting Governor Spencer Cox), https://www.nytimes.com/2025/09/14/opinion/charlie-kirk-shooting-internet.html [https://perma.cc/7U6P-77H9].

    Return to citation ^

  214. ^ Fyodor Dostoevsky, The Brothers Karamazov 237 (Richard Pevear & Larissa Volokhonsky trans., Farrar, Straus & Giroux paperback ed. 2002) (1880).

    Return to citation ^
Download
Topics:

February 10, 2026

More from this Issue

See Full Issue