Published on
Updated on
NOTE
Patterson v. Meta Platforms, Inc., 239 N.Y.S.3d 726 (App. Div. 2025)
Layton M. Moser*
I. Introduction
In May 2022, a teenage gunman drove over 200 miles to commit a racially motivated mass shooting at a Buffalo grocery store.[1] The shooter attributed his radicalization to extremist content, specifically the “Great Replacement Theory” that he was exposed to on technology platforms over many months.[2] In the tragedy’s aftermath, survivors and victims’ families filed civil actions against Meta, Google, YouTube, Reddit, and other platforms, but they faced a formidable legal barrier—Section 230 of the Communications Decency Act (“CDA”), which grants sweeping immunity to online platforms for third-party content.[3]
The Plaintiffs in Patterson v. Meta Platforms, Inc. attempted a relatively new legal strategy.[4] Rather than challenging what content the platforms published, they attacked how the platforms were designed.[5] They alleged that technology companies deliberately engineer addictive features, such as late-night push notifications, infinite scroll, autoplay videos, and gamification elements, all designed to maximize user engagement.[6] The Plaintiffs argued that these design defects caused the shooter’s addiction and radicalization, making Patterson a products liability case rather than a content moderation dispute.[7]
The New York Appellate Division disagreed.[8] The court dismissed all claims, holding that Section 230 immunity extends even to product design challenges when those designs involve organizing and displaying third-party content.[9] Yet two dissenting justices identified what the majority overlooked—a critical distinction between editorial functions and product design choices that should face ordinary tort liability.[10] Their dissent offers a framework for addressing the documented public health crisis of social media addiction while preserving Section 230’s important protections for content moderation.[11]
This Note explores the tension between Section 230 immunity and products liability law in the context of technology platforms’ addictive design features. Part II discusses the essential facts and procedural history of Patterson. Part III provides an overview of Section 230’s immunity framework, the evolving jurisprudence on algorithmic content curation, First Amendment protections for editorial functions, and traditional products liability principles. Part IV explains the New York Appellate Division’s holding and the dissenting justices’ alternative framework. Finally, Part V argues that the dissent correctly identified the core distinction between platforms’ roles as content publishers and product designers and demonstrates why courts should adopt this approach to preserve Section 230’s purpose while allowing accountability for deliberately addictive design features that harm vulnerable users.
II. Facts and Holding
After the teen shot and killed ten people and received life without the possibility of parole, a group of survivors and family members of the victims filed civil actions alleging various tort claims, including negligence, unjust enrichment, and strict products liability for defective design and failure to warn.[12] The Defendants are a group of technology platforms, including Meta Platforms, Inc. (Facebook); Instagram LLC; Snap Inc.; Alphabet, Inc.; Google, LLC; YouTube, LLC; Discord, Inc.; Reddit, Inc.; Twitch Interactive, Inc.; Amazon.com, Inc.; and 4chan Community Support, LLC.[13] According to the Plaintiffs, the technology platforms were defectively designed with content-recommendation algorithms that fed racist and violent content to the shooter, causing his radicalization over time.[14] Plaintiffs further alleged that the platforms were designed to be addictive through features like badges, streaks, trophies, and emojis, with the shooter’s addiction to these platforms ultimately influencing him to commit mass murder.[15] Plaintiffs asserted these claims against platforms using content-recommendation algorithms and those without such algorithms, arguing that even non-algorithmic platforms contained the same core defect of addictive design.[16]
The technology company Defendants moved to dismiss the claims, asserting immunity under Section 230 of the Communications Decency Act (“CDA”) and First Amendment protections.[17] The trial court denied the motion to dismiss in March 2024, stating that the Defendants should be held liable as product designers rather than publishers of third-party content.[18] The Defendants appealed this ruling, and the Appellate Division reversed the ruling of the trial court, dismissing all complaints against Defendants.[19]
III. Legal Background
The legal framework that governs technology platforms’ liability remains contested. As courts grapple with whether and when technology platforms can be held liable for harms caused by their design features and content-recommendation algorithms, they are weighing the future of innovation, free speech, and public safety. This Section explores the competing legal doctrines at the heart of this tension. First, Part A examines Section 230 of the CDA and the broad immunity it provides to platforms for third-party content. It analyzes the emerging circuit split over whether algorithmic content curation strips platforms of Section 230 protection or constitutes traditional editorial functions. It then discusses First Amendment protections for platforms’ expressive activity and the uncertain application of these protections in private tort liability. Finally, Part B outlines products liability principles and how courts have begun applying them to digital platforms, setting the stage for understanding the legal questions presented in Patterson.
A. Section 230 of the Communications Decency Act
The Communications Decency Act (“CDA”) was enacted by Congress in 1996 to “preserve the vibrant and competitive free market” for the Internet.[20] Congress sought to address the specific problem that early court decisions, which threatened to hold online platforms liable for third-party content if they engaged in any content moderation, creating a troublesome incentive to ignore harmful content entirely.[21] Section 230 was designed to solve this dilemma by immunizing platforms from liability for third-party content while simultaneously encouraging them to remove undesirable material.[22]
Section 230(c)(1) states that no provider of an interactive computer service shall be treated as the publisher or speaker of information provided by another information content provider.[23] This creates federal immunity to any cause of action that would make service providers liable for information originating with third-party users of the site.[24] Section 230(c)(2) offers additional protection for platforms’ good faith efforts to restrict access to undesired material, even if that material is constitutionally protected.[25] Federal courts have generally construed Section 230(c) broadly in favor of immunity, recognizing Congress’s intent to promote the continued development of the Internet free from federal or state regulation.[26]
To determine whether Section 230 immunity applies, courts apply a three-part test established by the Ninth Circuit in Barnes v. Yahoo!.[27] First, the defendant must be a provider or user of an interactive computer service.[28] Second, the plaintiff must seek to treat the defendant, under state law, as a publisher or speaker.[29] Third, the claim must involve information provided by another information content provider.[30] The critical inquiry typically centers on the second and third prongs of this test, whether the plaintiff’s theory of liability would treat the defendant as a publisher of third-party content, regardless of how the claim is labeled.[31] Courts focus not on the name given to the cause of action but “whether a plaintiff’s theory of liability would treat a defendant as a publisher or speaker of third-party content.”[32] This approach prevents plaintiffs from circumventing Section 230 merely by framing content-based claims as product defect claims.[33]
1. Algorithmic Curation and the Boundaries of Section 230 Protection
A key question in the modern Section 230 jurisprudence is whether content-recommendation algorithms strip away publisher immunity. The majority of circuits have held that algorithms matching content with user interests do not deprive platforms of Section 230 protection.[34] In Force v. Facebook, the Second Circuit held that using algorithms “designed to match that information with a consumer’s interest” does not make the platform the developer of third-party content.[35] The court reasoned that simply arranging others’ content in a certain way with these algorithms does not make platforms responsible as developers of that content, as doing otherwise would contravene Section 230’s text and purpose.[36]
Similarly, in M.P. v. Meta Platforms, the Fourth Circuit characterized content-recommendation algorithms as traditional editorial functions of publishers, comparing Facebook’s algorithm to a newspaper editor deciding where to place an article.[37] The Plaintiff alleged that Facebook’s algorithms radicalized a mass shooter by directing white supremacist content towards him.[38] The court held that “[d]ecisions about whether and how to display certain information provided by third parties are traditional editorial functions of publishers, notwithstanding the various methods they use in performing that task.”[39] Under this view, decisions about displaying third-party content constitute core publishing activity protected by Section 230, whether accomplished through human editors or automated algorithms.[40]
However, a circuit split has emerged on this issue. The Third Circuit adopted a minority position in Anderson v. TikTok, holding that algorithms can transform third-party content into first-party speech by constituting “expressive activity” under the First Amendment.[41] In Anderson, a ten-year-old died attempting a dangerous choking game that TikTok’s algorithm had recommended via a personalized “For You Page.”[42] The Third Circuit held that when TikTok recommends videos, it transforms user-created content into its own speech, because the act of curation is an “expressive activity.”[43] Since Section 230 only protects publication of third-party content, not a platform’s own speech, the court reasoned that TikTok could not claim immunity under the statute.[44] This reasoning suggests that platforms using content-recommendation algorithms could finally be held liable for the material they recommend, as the algorithmic curation itself morphs into the platform’s speech rather than mere republication of others’ content.[45]
This conflict between Anderson and the decisions in Force and M.P., creates a significant uncertainty for platforms operating in the United States about when their algorithmic content curation enjoys Section 230 protection.[46] If the approach in Anderson were to be widely adopted, platforms could face liability for each applicable tort claim arising from dangerous content recommended by their algorithms, which would fundamentally undermine the framework of Section 230.[47] Yet Anderson also highlights a real concern: highly sophisticated algorithms that curate personalized content feeds may involve editorial judgments that go well beyond passive republication of third-party material.[48]
2. First Amendment Protection for Algorithmic Curation
Separate from but related to Section 230 immunity, the First Amendment provides an additional layer of protection for platforms’ editorial choices.[49] In Moody v. NetChoice, the Supreme Court held that content-moderation algorithms constitute expressive activity protected by the First Amendment.[50] Moody involved challenges to Florida and Texas state laws that sought to prevent large technology platforms from engaging in what the states characterized as “censorship” of particular viewpoints.[51] Florida’s law prohibited platforms from permanently banning political candidates or journalistic enterprises, while Texas’s law barred platforms from removing content based on the viewpoints expressed.[52] Both laws essentially mandated that platforms carry and disseminate speech they might otherwise choose to exclude from their services.[53] The Court explained that “deciding on the third-party speech that will be included in or excluded from a compilation—and then organizing and presenting the included terms—is expressive activity of its own.”[54] Just like a newspaper editor choosing where to place articles, technology platforms engage in protected expression when they organize third-party content.[55]
However, the Moody Court noted that its holding did not address “feeds whose algorithms respond solely to how users act online—giving them the content they appear to want, without any regard to independent content standards.”[56] The Court’s analysis focused on government restrictions on speech, not private tort liability between parties.[57] The application of the Moody First Amendment framework to products liability claims remains uncertain and presents the critical question of whether tort claims seeking reasonable design safeguards for addictive features should trigger the same First Amendment scrutiny as direct government censorship. The answer to this question has profound implications for whether victims of platform-related harms can seek any legal remedy, or whether the First Amendment constructs a barrier to accountability for design choices that cause legitimate injuries.[58]
B. Products Liability Principles and Their Application to Technology Platforms
Traditional products liability law imposes strict liability for defective products, holding that manufacturers who place products into the stream of commerce bear responsibility for injuries caused by defects, regardless of the care exercised.[59] Courts recognize three types of defects: (1) manufacturing defects which are departures from the intended design, (2) design defects which are foreseeable risks that could be avoided by a reasonable alternative design, and (3) failure to warn, which includes inadequate instructions or warnings about foreseeable risks.[60] For design defects, plaintiffs are required to show that “the foreseeable risks of harm posed by the product could have been reduced or avoided by the adoption of a reasonable alternative design . . . and the omission of the alternative design renders the product not reasonably safe.”[61]
The application of products liability law to technology platforms remains contested and ever-changing. A threshold question is whether platforms constitute “products” at all.[62] The Third Restatement defines products as “tangible personal property distributed commercially for use or consumption,” but recognizes that intangible items can also be products when distributed and used similarly to tangible property.[63] Courts have applied this approach and found that software, computer programs, and digital services can constitute products in appropriate circumstances.[64]
Plaintiffs have increasingly attempted to characterize technology platforms as products, which would be subject to strict liability rather than the “services” protected by Section 230.[65] This distinction matters because Section 230 does not apply when harm flows directly from a platform’s own design features unrelated to third-party content.[66] The critical question becomes whether the alleged defect involves the platform’s role as a publisher of third-party content or instead involves design features that are outside of that publishing function.[67]
The Ninth Circuit’s decision in Lemmon v. Snap illustrates this principle.[68] There, plaintiffs alleged that Snapchat’s “Speed Filter,” which displayed users’ real-time speed while driving, encouraged reckless driving by incentivizing users to record their face while speeding, ultimately leading to a fatal crash.[69] The Court held that the “Speed Filter” was Snapchat’s own created feature rather than third-party content, and thus fell outside Section 230’s protection.[70] The Court emphasized that the “duty to design a reasonably safe product is fully independent of [a defendant’s] role in monitoring or publishing third-party content.”[71] Since plaintiffs challenged a feature that Snapchat itself created—not content posted by users—the products liability claim could proceed.[72] However, the Court was careful to note the limits of this exception. Plaintiffs “would not be permitted under § 230(c)(1) to fault Snap for publishing other Snapchat-user content . . . that may have incentivized . . . dangerous behavior.”[73] The distinction between a platform’s own design features and its role in facilitating third-party speech creates a seemingly simple framework. If the alleged harm flows from third-party content, Section 230 bars liability.[74] If the harm flows from the platform’s own features unrelated to content, products liability law kicks in and strict liability may apply.[75]
More recently, district courts have grappled with whether these algorithmic content-recommendation systems constitute products that are subject to strict liability claims.[76] In A.M. v. Omegle.com, the court allowed products liability claims to proceed where the plaintiff challenged the platform’s matching algorithm that paired minors with adults for video chats.[77] The Plaintiff alleged that Omegle’s design was defective for failing to implement age verification or matching restrictions.[78] The court held that Section 230 did not bar these claims because the Plaintiff did not seek to have Omegle review, edit, or remove any third-party content.[79] Rather, the Plaintiff challenged design choices unrelated to content moderation, arguing that Omegle could have implemented safeguards in how the platforms matched users without touching the substance of their communications.[80]
Against this backdrop of evolving Section 230 jurisprudence, an emerging circuit split on algorithmic curation, First Amendment protections for editorial functions, and developing products liability theories, a New York appellate court confronted Plaintiffs’ claims that technology platforms’ addictive design features and algorithmic content curation caused the Buffalo shooter’s radicalization and the following mass murder.[81]
IV. Instant Decision
Much like the many aforementioned cases, the court in Patterson had to work through the legal tangle of whether Section 230 applied and whether the social media algorithms at issue could be considered products, subject to strict liability.[82] The appellate court reversed the trial court’s ruling and dismissed all claims against the technology company Defendants, holding that Section 230 of the CDA barred the Plaintiffs’ tort claims, even when framed as a product liability suit.[83]
The court applied the three-part test from Barnes, finding that Plaintiffs sought to treat Defendants as publishers of third-party content.[84] Further, the court rejected Plaintiffs’ argument that they sued the Defendants as product designers, reasoning that the claims were “inextricably intertwined” with the Defendants’ publishing role.[85] Backing its reasoning with Force and M.P., the majority held that content-recommendation algorithms constitute traditional editorial functions that do not strip away Section 230 protection.[86] The court noted that the Plaintiffs’ theory of harm depended on the racist content the shooter consumed rather than on platform design alone, which is precisely the type of content-based claim that Section 230 bars.[87] The opinion held additionally that algorithmic content curation constitutes expressive activity protected by the First Amendment under Moody, creating what the court called a “Heads I win, tails you lose” situation where platforms receive either Section 230 immunity or First Amendment protection.[88]
However, Justices Bannister and Nowak dissented, arguing that the majority misread the complaints.[89] The duo of dissenters believed the Plaintiffs alleged that platforms are defectively designed because of deliberate choices that make them addictive—not because of third-party content.[90] The dissent detailed how platforms prey on young users through push notifications at night, autoplay and infinite scroll features, and inadequate parental controls.[91] These allegations seek accountability for failing to provide basic safety features, not for publishing content.[92]
Justices Bannister and Nowak concluded that platforms qualify as products under New York law because the Defendants control the design, are positioned to prevent harms, and distribute uniform products.[93] Citing to Lemmon v. Snap, the dissent emphasized that the duty to design safe products is independent of any publishing function.[94] Their opinion rejected the majority’s First Amendment analysis, noting that Moody involved government censorship rather than private tort liability, and warned against expanding protections so broadly that no tort liability could exist for design defects.[95] Finally, the two judges offered a framework distinguishing content-based claims from design-based products liability, allowing accountability for harmful design while preserving Section 230’s protection for editorial decisions.[96]
V. Comment
Technology platforms have fundamentally transformed how the average American communicates, consumes information, and interacts with others. As of 2024, approximately 72% of American adults use at least one social media platform,[97] with the average user spending over two hours per day on these services.[98] Among teenagers aged thirteen to seventeen, usage rates reach as high as 95%, with over one-third of these young users reporting near-constant engagement with social media throughout their day.[99] Despite platforms typically requiring users to be at least thirteen years old, approximately two in five children between the ages of eight and twelve actively use social media services.[100] The economic stakes here are equally significant, as the social media industry has generated over $150 billion in revenue in 2025, driven mainly by advertising models that depend on maximizing user engagement.[101] Yet this widespread adoption has come with noted costs. Recent studies have linked social media use to increased rates of anxiety, depression, and addiction—particularly among adolescents.[102] More than one in ten adolescents show signs of problematic social media behavior, in which they struggle to control their usage despite negative consequences.[103] For platforms whose business models depend on keeping users engaged, these noted harms raise critical questions about legal accountability.[104]
The appellate court’s decision to dismiss Plaintiffs’ claims in Patterson is not surprising given the current trajectory of Section 230 jurisprudence and the broad immunity that federal courts have afforded to technology platforms. Yet the clear pattern—and problem—emerging from Patterson is the judiciary’s reluctance to distinguish between platforms’ roles as content publishers and their roles as product designers. The result is a legal framework that grants near-absolute immunity to billion-dollar technology companies for design choices that cause legitimate harms, particularly to vulnerable minors.
This problem, however, is not impossible to overcome. The dissent in Patterson offers a workable solution: treating content moderation and addictive design features as separable functions subject to different liability rules.[105] Technology platforms require protection for their editorial judgments on third-party speech, and Section 230 provides this. In an era where social media is one of the most popular vehicles for information and communication, some protection is absolutely necessary. However, platforms should not receive blanket immunity for deliberately engineering addictive features that exploit the psychological vulnerabilities of adolescents. The real problem lies in the majority’s failure to recognize this critical distinction, collapsing all functions into a single “publishing” category that Section 230 was never intended to protect so completely.
Part A demonstrates that the Patterson dissent correctly identified a core issue by recognizing that Plaintiffs challenged addictive product design features rather than third-party content publication, supported by scientific evidence of deliberate engineering choices that exploit psychological vulnerabilities. Part B then argues that the dissent’s Section 230 and First Amendment analyses are sound. Highly personalized algorithmic curation constitutes content development rather than mere republication; extending First Amendment protections from government censorship to private tort liability would create absurd results inconsistent with centuries of tort law. Part C argues that policy considerations overwhelmingly favor the dissent’s approach. This approach preserves Section 230’s core purpose of encouraging content moderation while still allowing accountability for addictive design features, creating better incentives for safer innovation without the downsides of blanket immunity. Finally, Part D concludes by demonstrating that the dissent’s framework offers a workable and necessary distinction between content moderation and product design, which also addresses mounting public health concerns while maintaining Section 230’s viability in the face of growing legislative and public pressure for reform.
A. The Dissent Correctly Identified the Core Issue: Product Design, Not Content Publication
The dissenting justices properly understood what the majority obscured: that the Plaintiffs’ claims were challenging how technology platforms are designed to be addictive, not what content they publish.[106] The complaints alleged specific design defects unrelated to third-party speech.[107] Platforms send push notifications throughout the night, prompting children to re-engage when they would otherwise be sleeping.[108] They employ autoplay video features and infinite-scroll capabilities that create constant streams of media which are difficult to close.[109] They lack adequate age verification tools and parental controls.[110] They intentionally place obstacles to discourage account deactivation or deletion.[111] None of these allegations require platforms to review, edit, or remove any third-party content.[112]
The dissent emphasized that Plaintiffs proposed reasonable alternative designs independent of content moderation, such as eliminating autoplay features, creating feed endpoints, providing time limits, implementing parental controls and age verification, and removing barriers to account deletion, to name a few.[113] Each of these design changes would address addictive architecture without touching any third-party content on the platforms.[114] This is precisely the distinction the majority failed to notice.[115] Moreover, these proposed design changes are both feasible and inexpensive to implement. Disabling autoplay, for instance, is largely a matter of changing a default setting—a simple adjustment shown to reduce user engagement time.[116] Likewise, adding time limits, parental controls, or easier account deletion involves modifying existing interface settings rather than overhauling content systems.[117] Building an entire social-media platform can cost tens or hundreds of thousands of dollars; small user experience updates like these only require a fraction of that investment.[118] In short, these fixes are technically simple and economically well within reach for these major platforms.
The scientific evidence supports the dissent’s characterizations. Roughly one in five teens state that social media sites hurt their mental health.[119] Young users aged eighteen to twenty-two account for 40% of all Americans addicted to social media.[120] The mental health consequences are severe: 48% of heavy users exhibit depression, while about 22% suffer from anxiety.[121] Teenagers spending five to seven hours daily on social media are twice as likely to exhibit mental health problems.[122] While the specific content users encounter certainly plays a role in these harms, the addictive features that keep users scrolling for several hours daily represent an independent and additional source of injury. Even if every piece of content were wholesome, the psychological mechanisms of the addiction would still inflict harm. When harmful content is delivered through an addictive product, the design features act as gasoline does on a fire, ensuring prolonged and repeated exposure that exacerbates the damage.
These statistics are a consequence of deliberate design decisions. In 2022, consumer social media apps generated over $167 billion, largely through these advertising models that are dependent on maximizing engagement.[123] Engineers and psychologists alike are employed by technology companies specifically to make products more addictive, testing features to determine what keeps users scrolling the longest.[124] They deploy variable reward schedules—the same psychological mechanism that makes slot machines addictive—to keep users checking feeds compulsively.[125] Platforms know their products exploit psychological vulnerabilities, particularly in adolescents whose brains are still developing, yet they deploy these features because their business models depend on maximizing engagement time.[126]
The dissent properly recognized that the Defendants are billion-dollar corporations uniquely positioned to prevent harms from social media addiction.[127] They alone control platform design and distribution.[128] Their platforms are uniform for all users.[129] The fact that users pay with data instead of dollars does not strip these platforms of their nature as commercial products.[130] Under New York law, the question of whether something is a product is intertwined with a duty to warn, with key factors including control over design, standardization, and superior ability to know about inherent dangers.[131] All of these factors point toward treating technology platforms as products subject to traditional liability principles.[132]
B. The Dissent’s Section 230 and First Amendment Analyses Are Sound
The dissent offered statutory and constitutional analysis that provides alternative pathways to accountability.[133] Section 230(f)(3) defines an information content provider as any person responsible for the “creation or development of information.”[134] If a defendant is a content provider, it receives no immunity.[135] The dissenters noted that algorithmic functions like autoplay and infinite feeds constitute creation or development of information, rendering the Defendants as first-party content providers.[136] This interpretation is also supported in Justice Katzmann’s concurrence in Force, which reasoned that “it strains the English language” to say platforms targeting specific content to specific users act merely as publishers rather than developers.[137]
The distinction is critical. Traditional publishers make editorial decisions about what to publish, but they publish the same content for all readers.[138] Social media algorithms create individualized content streams tailored to maximize each user’s engagement.[139] Some users receive cooking videos; others receive white nationalist content.[140] Neither group knows what the other sees.[141] This is not traditional publishing; it is active content development driven by learning systems designed to determine what content will maximize user engagement.[142] When an algorithm determines that showing a user extremist content will maximize engagement and then funnels that content to them—while showing others different content—the platform has moved beyond republishing into active development of personalized information products.[143]
The dissent also correctly rejected the majority’s First Amendment analysis. The majority relied on the Moody decision, but Moody involved government-imposed content restrictions—direct state censorship—not private tort liability.[144] Extending Moody to also protect platforms from products liability represents an inequitable expansion of the First Amendment’s application to this issue.[145]
The dissenters warned of logical problems flowing from the majority’s reasoning.[146] Under their logic, there could never be tort liability for failing to warn of product risks, as warnings constitute compelled speech.[147] There could never be medical malpractice liability for lack of informed consent, as explaining procedures implicates speech.[148] Cigarette warning labels would be unconstitutional. These ridiculous results display the flawed reasoning in the majority’s approach.
The distinction between government censorship and private tort liability matters enormously. When government prohibits speech, it uses sovereign power to restrict the marketplace of ideas, the First Amendment’s core concern.[149] When a private plaintiff sues for injuries from defective design, no state censorship occurs.[150] The platform remains free to organize content however it wishes, subject only to potential liability if its design causes foreseeable harm.[151] This is ordinary tort law, which has coexisted with the First Amendment for over two centuries.
C. Policy Considerations Favor the Dissent’s Approach
Policy considerations overwhelmingly favor distinguishing design accountability from content regulation.[152] The dissent’s framework preserves Section 230’s core purpose while allowing necessary accountability for design choices causing documented public health harms.[153]
Section 230 was enacted to solve the “moderator’s dilemma,” which stemmed from early decisions that threatened to hold platforms liable for third-party content if they moderated at all.[154] The dissent’s framework preserves this function of Section 230. Platforms remain free to make editorial judgments about third-party content. They can use algorithms to organize and present content based on editorial standards. They can prioritize topics, demote misinformation, and remove hate speech—all without liability. Section 230 continues protecting these publishing functions. The dissent’s approach only exposes platforms to potential liability for addictive design features, especially those targeting vulnerable minors.[155]
This distinction maintains appropriate incentives. Platforms should be encouraged to moderate harmful content, and Section 230 ensures this. However, platforms should not wield the ability to addict children through manipulative design. Products liability discourages such choices.[156] This framework aligns legal incentives with social welfare. The evidence of harm is overwhelming and growing, as teenagers who use social media for three or more hours daily face an increased risk of depression and anxiety, yet teens now average 4.8 hours on these platforms daily.[157] Nearly half of all teens admit they spend too much time on social media,[158] and over half of all teens believe it would be very or somewhat hard to give up social media.[159] The mental health consequences are staggering, with 40% of youth students experiencing persistent feelings of sadness and hopelessness and more than 20% seriously considering suicide.[160]
These harms are not accidental; they are the result of deliberate design choices. Former Facebook executive Sean Parker admitted the platform was designed to exploit a “vulnerability in human psychology” by giving users a dopamine hit via a social validation loop.[161] Parker said that the inventors knew about this and did it anyway.[162] Platforms intentionally design features that overstimulate the brain’s reward center and trigger pathways comparable to addiction, particularly in adolescent brains.[163] Internal documents from Meta reveal that the company’s own research found that Instagram makes body image issues worse for one in three teenagers, yet executives chose not to act on these findings.[164] Former U.S. Surgeon General Dr. Vivek Murthy emphasized that adolescent brain development makes young people uniquely vulnerable to these addictive design features, noting how difficult it becomes for teenagers to disengage when platforms employ sophisticated techniques specifically designed to maximize their time on the platform.[165] He characterized the situation as placing teenagers in an inherently unfair competition against the world’s best product engineers, who leverage cutting-edge neuroscience to increase user engagement.[166]
The majority’s approach creates perverse incentives by granting blanket immunity for all design decisions that fall under “publishing,” effectively eliminating any incentive for safe designs.[167] Why would platforms implement parental controls, time limits, or age verification when their business models depend on maximizing engagement, and they face no legal consequences for addicting their users? The majority encourages these platforms to make products as addictive as possible because only in very rare situations would they face liability.[168] Yet their warnings about the “end of the Internet” are overstated.[169] Other industries, such as pharmaceuticals and automobiles, innovate successfully while adhering to safety standards. Technology platforms can operate under the same regime, using algorithms to maximize revenue, while at the same time implementing reasonable safeguards for vulnerable users without stifling innovation.
The dissent’s approach may prove necessary to preserve Section 230 long-term. With over 64% of Americans believing social media has a net negative effect and more than forty state attorneys general suing platforms for deliberately addicting children,[170] public pressure is mounting.[171] New York City’s health commissioner officially declared social media a public health hazard in 2024 and filed a federal lawsuit alleging platforms created a public nuisance by intentionally exploiting the psychology of young users.[172] Over 2,050 similar lawsuits are now pending nationwide, representing governments, school districts, and individual plaintiffs.[173] Without a nuanced middle path, legislative backlash may eliminate Section 230’s protections entirely, as over a hundred bills to regulate technology platforms were introduced by state legislatures in 2024 alone.[174]
Justices Bannister and Nowak offer a nuanced approach that protects content moderation while allowing design accountability. They offer a middle path that addresses public concerns while preserving the all-important benefits of Section 230. The dissent also provides a much clearer approach than the majority’s “inextricably intertwined” standard. The workable test is that if the proposed remedy requires reviewing, editing, or removing any third-party content, Section 230 applies; if the remedy involves design changes that can be implemented without touching any content, product liability may follow. This clear distinction provides practicability for all parties while maintaining appropriate incentives for safer innovation.
VI. Conclusion
Ten people died in that Buffalo grocery store: mothers, fathers, grandparents, and community members shopping on a Saturday afternoon. The shooter drove over 200 miles to commit his racially motivated massacre after spending months on technology platforms being fed extremist content through algorithms designed to maximize his engagement. When the victims’ families sought accountability, they did not challenge what content the platforms published—they challenged how the platforms were deliberately designed to be addictive. The majority refused to see this distinction, granting blanket immunity under a statute never intended to shield product design choices.
The dissent recognized what the majority obscured: that platforms deserve protection for editorial decisions about content, but they should face ordinary tort liability for engineering addictive features that exploit psychological vulnerabilities. Late-night push notifications, infinite scroll, autoplay videos, variable reward schedules borrowed from slot machines—these design choices prey on developing brains, particularly adolescents. Yet the current legal landscape offers victims no remedy and creates no incentive for safer design, a heads I win, tails you lose situation. The question is not whether platforms deserve protection for editorial functions—they do. The question is whether billion-dollar corporations should receive absolute immunity for deliberately addictive designs that cause documented harm to vulnerable users. The Patterson dissent answered correctly—they should not.
*B.A. Political Science, University of Missouri, 2024; J.D. Candidate, University of Missouri School of Law, 2027; Associate Member, Missouri Law Review, 2025–2026. I am thankful to Professor Renee Henson and her professional guidance throughout the writing of this note, as well as the Missouri Law Review, Caleb Ross, Thomas Yang, Claire Carlson, and Andrew Bohon for their assistance in the editing process.
[1] Patterson v. Meta Platforms, Inc., 239 N.Y.S.3d 726, 728 (App. Div. 2025).
[2] Id. The “Great Replacement Theory” refers to a hypothetical plot to decrease the influence of white people, specifically in the United States. See David Bauder, What is ‘great replacement theory’ and how does it fuel racist violence?, PBS News (May 16, 2022), https://www.pbs.org/newshour/politics/what-is-great-replacement-theory-and-how-does-it-fuel-racist-violence.
[3] Id. at 729. Note that third party content refers to any material or data created by an external individual or organization, rather than by the company operating the platform. See 47 U.S.C. § 230.
[4] See Patterson, 239 N.Y.S.3d at 729.
[5] Id.
[6] Id.
[7] Id.
[8] Id. at 730.
[9] Id.
[10] Id. at 736 (Bannister & Nowak, JJ., dissenting).
[11] Id.
[12] Id. at 728–29.
[13] Id. at 729.
[14] Id.
[15] Id.
[16] Id.
[17] Id. at 730.
[18] Id.
[19] Id.
[20] 47 U.S.C. § 230(b)(2).
[21] Patterson, 239 N.Y.S.3d at 730.
[22] Id.
[23] Hon. John G. Browning, A Product by Any Other Name? The Evolving Trend of Product Liability Exposure for Technology Platforms, 16 Elon L. Rev. 181, 187 (2024).
[24] Id. at 186.
[25] See 47 U.S.C. § 230(c)(2).
[26] Force v. Facebook, Inc., 934 F.3d 53, 64 (2d Cir. 2019).
[27] Barnes v. Yahoo!, Inc., 570 F.3d 1096, 1100–01 (9th Cir. 2009).
[28] Id. at 1100.
[29] Id.
[30] Id. at 1101.
[31] Id.
[32] Patterson v. Meta Platforms, Inc., 239 N.Y.S.3d 726, 731 (App. Div. 2025).
[33] Browning, supra note 23, at 188.
[34] See generally Force v. Facebook, Inc., 934 F.3d 53 (2d Cir. 2019) (holding that algorithms that match content with user interests do not make platforms lose Section 230 protection).
[35] Id. at 66.
[36] Id. at 70.
[37] M.P. ex rel. Pinckney v. Meta Platforms, Inc., 127 F.4th 516, 526 (4th Cir. 2025).
[38] Id. at 521.
[39] Id. at 526.
[40] Id.
[41] Anderson v. TikTok, Inc., 116 F.4th 180, 184 (3d Cir. 2024).
[42] Id. at 182.
[43] Id. at 184.
[44] Id.
[45] See id.
[46] See, e.g., Carafano v. Metrosplash.com, Inc., 339 F.3d 1119, 1124 (9th Cir. 2003).
[47] Patterson v. Meta Platforms, Inc., 239 N.Y.S.3d 726, 732–33 (App. Div. 2025).
[48] See Anderson, 116 F.4th at 184.
[49] Patterson, 239 N.Y.S.3d at 732.
[50] Moody v. NetChoice, LLC, 603 U.S. 707, 731–32 (2024).
[51] Id. at 707.
[52] Id. at 720–21.
[53] See id.
[54] Id. at 731.
[55] Id.
[56] Id. at 736 n.5.
[57] Id. at 733–34.
[58] See Patterson v. Meta Platforms, Inc., 239 N.Y.S.3d 726 (App. Div. 2025).
[59] Browning, supra note 23, at 184.
[60] Id. at 185.
[61] Restatement (Third) Of Torts: Prods. Liab. § 2(b) (A.L.I. 1998).
[62] Browning, supra note 23, at 202.
[63] Restatement (Third) Of Torts: Prods. Liab. § 19(a) (A.L.I. 1998).
[64] See Anderson v. TikTok, Inc., 116 F.4th 180 (3d Cir. 2024); Lemmon v. Snap, Inc., 995 F.3d 1085 (9th Cir. 2021).
[65] See In re Soc. Media Adolescent Addiction/Pers. Inj. Prods. Liab. Litig., 702 F. Supp. 3d 809 (N.D. Cal. 2023).
[66] See, e.g., Patterson v. Meta Platforms, Inc., 239 N.Y.S.3d 726 (App. Div. 2025). Plaintiffs attempted to characterize AI algorithms as “products” not immune from Section 230 liability. Id.
[67] Lemmon, 995 F.3d at 1091.
[68] Id.
[69] Id. at 1091–92.
[70] Id. at 1094.
[71] Id. at 1093.
[72] Id. at 1094.
[73] Id. at 1093 n.4.
[74] See id.
[75] See id.
[76] See A.M. v. Omegle.com, LLC, 614 F. Supp.3d 814, 817 (D. Or. 2022).
[77] Id. at 818.
[78] Id. at 820.
[79] Id. at 819.
[80] Id.
[81] Patterson v. Meta Platforms, Inc., 239 N.Y.S.3d 726 (App. Div. 4th Dep’t. 2025).
[82] Id. at 729.
[83] Id. at 736.
[84] Id. at 731.
[85] Id. at 735.
[86] Id. at 733–34 (citing Force v. Facebook, Inc., 934 F.3d 53, 71 (2d Cir. 2019); M.P. ex rel. Pinckney v. Meta Platforms, Inc., 127 F.4th 516, 527 (4th Cir. 2025)).
[87] Patterson, 239 N.Y.S.3d at 735.
[88] Id. at 734.
[89] Id. at 737 (Bannister & Nowak, JJ., dissenting).
[90] Id.
[91] Id.
[92] Id. at 738.
[93] Id. at 740.
[94] Id. at 738.
[95] Id. at 741.
[96] Id. at 740.
[97] Social Media Fact Sheet, Pew Rsch. Ctr. (Nov. 13, 2024), https://www.pewresearch.org/internet/fact-sheet/social-media/.
[98] Simon Kemp, Digital 2024: Global Overview Report, DataReportal (Jan. 31, 2024), https://datareportal.com/reports/digital-2024-global-overview-report.
[99] The U.S. Surgeon General’s Advisory, Social Media and Youth Mental Health, Nat’l Libr. of Med.: Nat’l Ctr. for Biotechnology Info. (2023), https://www.ncbi.nlm.nih.gov/books/NBK594759/.
[100] Id.
[101] Social Networking – Worldwide, Statista, https://www.statista.com/outlook/amo/app/social-networking/worldwide?srsltid=AfmBOop7SumTzYKox-T7LPMPLVDkxohwKg1lhima9eOQ-xGmKm0-Qy6M (last visited Oct. 17, 2025).
[102] Holly Shannon et al., Problematic Social Media Use in Adolescents and Young Adults: Systematic Review and Meta–analysis, JMIR Publ’ns (Apr. 14, 2022), https://mental.jmir.org/2022/4/e33450?utm_source=chatgpt.com.
[103] Id.
[104] Browning, supra note 23, at 187.
[105] See Patterson v. Meta Platforms, Inc., 239 N.Y.S.3d 726, 737 (App. Div. 2025) (Bannister & Nowak, JJ., dissenting).
[106] Id.
[107] Id.
[108] Id.
[109] Id.
[110] Id. See also Kyooeun Jang, Lulia Pan & Nicol Turner Lee, The fragmentation of online child safety regulations, Brookings (Aug. 14, 2023), https://www.brookings.edu/articles/patchwork-protection-of-minors/.
[111] Patterson, 239 N.Y.S.3d at 737 (Bannister & Nowak, JJ., dissenting).
[112] Id. at 739.
[113] Id. at 737–38.
[114] Id. at 738.
[115] Id. at 726.
[116] Miranda Redenbaugh, UChicago scientists study the hidden cost of Netflix’s autoplay, Univ. of Chi.: UChicago News (Mar. 3, 2025), https://news.uchicago.edu/story/uchicago-scientists-study-hidden-cost-netflixs-autoplay.
[117] Vesselina Lezginov, Breaking Down Software Development Costs: A Complete Guide, Scopic (Nov. 8, 2024), https://scopicsoftware.com/blog/software-development-costs/.
[118] Id.
[119] Michelle Faverio, Monica Anderson & Eugenie Park, Teens, Social Media and Mental Health, Pew Rsch. Ctr. (Apr. 22, 2025), https://www.pewresearch.org/internet/2025/04/22/teens-social-media-and-mental-health/.
[120] Jessica Miller, Social Media Addiction Statistics, AddictionHelp.com (Oct. 19, 2025), https://www.addictionhelp.com/social-media-addiction/statistics/.
[121] Ujala Zubair, Muhammad K. Khan & Muna Albashari, Link between excessive social media use and psychiatric disorders, Nat’l Libr. Of Med.: Nat’l Ctr. for Biotechnology Info. (Mar. 27, 2023), https://pmc.ncbi.nlm.nih.gov/articles/PMC10129173/.
[122] Social Media Addiction Statistics, Lanier Law Firm, https://www.lanierlawfirm.com/social-media-addiction/statistics/ (last visited Apr. 24, 2026).
[123] Browning, supra note 23, at 182.
[124] Alejandro L. Mujica et al., Addiction By Design: Some Dimensions and Challenges of Excessive Social Media Use, Med. Res. Archives, Feb. 24, 2022, at 3.
[125] Daniel Kruger, Social media copies gambling methods ‘to create psychological cravings,’ Univ. of Mich.: Inst. for Healthcare Pol’y & Innovation (May 8, 2018), https://ihpi.umich.edu/news/social-media-copies-gambling-methods-create-psychological-cravings?utm_source=chatgpt.com.
[126] Mujica et al., supra note 124, at 3.
[127] Patterson v. Meta Platforms, Inc., 239 N.Y.S.3d 726, 740 (App. Div. 2025) (Bannister & Nowak, JJ., dissenting).
[128] Id.
[129] Id.
[130] Id.
[131] Id. at 739.
[132] Id. at 740.
[133] Id. at 737.
[134] 47 U.S.C. § 230(f)(3).
[135] Patterson, 239 N.Y.S.3d at 740 (Bannister & Nowak, JJ., dissenting).
[136] Id.
[137] Force v. Facebook, Inc., 934 F.3d 53, 76–77 (2d Cir. 2019) (Katzmann, J., concurring).
[138] Patterson, 239 N.Y.S.3d at 741 (Bannister & Nowak, JJ., dissenting).
[139] Id.
[140] Id.
[141] Id.
[142] Id.
[143] Id.
[144] Id.
[145] Id.
[146] Id.
[147] Id. Compelled speech is a First Amendment doctrine prohibiting the government from forcing individuals to express messages or beliefs they do not hold. W. Va. State Bd. of Educ. v. Barnette, 319 U.S. 624, 641 (1943).
[148] Patterson, 239 N.Y.S.3d at 741 (Bannister & Nowak, JJ., dissenting).
[149] See Abrams v. United States, 250 U.S. 616, 630 (1919) (Holmes, J., dissenting).
[150] See Patterson, 239 N.Y.S.3d at 741 (Bannister & Nowak, JJ., dissenting).
[151] Id. at 741–42.
[152] See id. at 739–40.
[153] Id. at 741.
[154] See generally Stratton Oakmont, Inc. v. Prodigy Servs. Co., No. 31063/94, 1995 WL 323710 (N.Y. May 24, 1995) (cited in many modern decisions as the case that began the dilemma that prompted Section 230’s creation).
[155] Patterson, 239 N.Y.S.3d at 740 (Bannister & Nowak, JJ., dissenting).
[156] See Hernandez ex rel. Emeterio v. Tokai Corp., 189 F.3d 489 (5th Cir. 1999); Braswell v. Cincinnati, Inc., 731 F.3d 1081 (10th Cir. 2013);Duford v. Sears, Roebuck & Co., 833 F.2d 407 (1st Cir. 1987).
[157] Peter Susic, 18+ Teen & Kids Screen Time Statistics (2024): Avg. Screen Time for Teens, Headphones Addict (Mar. 28, 2024) https://headphonesaddict.com/teen-kids-screen-time-statistics/; Jonathan Rothwell, Teens Spend Average of 4.8 Hours on Social Media Per Day, Gallup (Oct. 13, 2023), https://news.gallup.com/poll/512576/teens-spend-average-hours-social-media-per-day.aspx.
[158] Faverio, Anderson & Park, supra note 119.
[159] Emily A. Vogels & Risa Gelles-Watnick, Teens and social media: Key findings from Pew Research Center surveys, Pew Rsch. Ctr. (Apr. 24, 2023), https://www.pewresearch.org/short-reads/2023/04/24/teens-and-social-media-key-findings-from-pew-research-center-surveys/.
[160] Social Media Addiction Statistics, supra note 122.
[161] Erica Pandey, Sean Parker: Facebook was designed to exploit human “vulnerability,” Axios (Nov. 9, 2017), https://www.axios.com/2017/11/09/sean-parker-facebook-was-designed-to-exploit-human-vulnerability-1513306782.
[162] Id.
[163] Kathy Katella, How Social Media Affects Your Teen’s Mental Health: A Parent’s Guide, Yale Med. (June 17, 2024), https://www.yalemedicine.org/news/social-media-teen-mental-health-a-parents-guide.
[164] Georgia Wells, Jeff Horwitz & Deepa Seetharaman, Facebook Knows Instagram Is Toxic for Teen Girls, Company Documents Show, Wall St. J. (Sept. 14, 2021), https://www.wsj.com/tech/personal-tech/facebook-knows-instagram-is-toxic-for-teen-girls-company-documents-show-11631620739.
[165] Kathryn Fink, Courtney Dorning & Mary Louise Kelly, U.S. surgeon general calls for tobacco-style warning labels for social media, NPR (June 17, 2024, at 17:35 ET), https://www.npr.org/2024/06/17/nx-s1-5008816/u-s-surgeon-general-calls-for-tobacco-style-warning-labels-for-social-media.
[166] Id.
[167] Patterson v. Meta Platforms, Inc., 239 N.Y.S.3d 726, 734–35 (App. Div. 2025).
[168] See id.
[169] See id. at 735.
[170] Brooke Auxier, 64% of Americans say social media have a mostly negative effect on the way things are going in the U.S. today, Pew Rsch. Ctr. (Oct. 15, 2020), https://www.pewresearch.org/short-reads/2020/10/15/64-of-americans-say-social-media-have-a-mostly-negative-effect-on-the-way-things-are-going-in-the-u-s-today/; Barbara Ortutay, More than 40 states sue Meta claiming its social platforms are addictive and harm children’s mental health, PBS News (Oct. 24, 2023, at 14:41 EDT), https://www.pbs.org/newshour/politics/more-than-40-states-sue-meta-claiming-its-social-platforms-are-addictive-and-harm-childrens-mental-health.
[171] See, e.g., In re Soc. Media Adolescent Addiction/Pers. Inj. Prods. Liab. Litig., 777 F. Supp. 3d 1016 (N.D. Cal. 2025); State v. Meta Platforms, Inc., 346 A.3d 489 (Vt. 2025); State ex rel. Jackson v. TikTok Inc., No. 24CV032063-910, 2025 WL 2399525 (N.C. Super. Ct. Aug. 19, 2025).
[172] New York City sues social media companies for allegedly addicting children, CNBC (Oct. 9, 2025, at 21:57 EDT), https://www.cnbc.com/2025/10/10/new-york-city-sues-social-media-companies-for-allegedly-addicting-children.html.
[173] Id.
[174] Rebecca Kern, Push to rein in social media sweeps the states, Politico (July 1, 2022, at 4:30 EDT), https://www.politico.com/news/2022/07/01/social-media-sweeps-the-states-00043229.