Last updated 6 months ago. Our resources are updated regularly but please keep in mind that links, programs, policies, and contact information do change.
A 19-year-old identified in court documents as K.G.M. took the stand this week in Los Angeles County Superior Court to describe how Instagram consumed her teenage years. Her testimony—part of a jury trial alleging that social media platforms deliberately designed addictive features that harmed children—arrives at a moment when Congress faces a question with no easy answer. Could lawmakers ban the design features that make apps so compulsive, and would the Constitution even allow it?
The Supreme Court ruled in 2024 that when platforms decide what content to display and how to arrange it—including through algorithms—the courts say platforms have the right to decide what to show. This creates an unusual alignment of interests: the platforms that designed these features and the Constitution that protects free expression both stand between lawmakers and the regulation that child safety advocates say is needed.
What Makes This Trial Different
The Los Angeles case represents one of the most significant instances of social media companies defending themselves before a jury on claims that their platform design itself—not the content users post—caused documented psychological harm to children. K.G.M.’s attorney, Matthew Bergman, previously represented plaintiffs in tobacco litigation. The parallel is intentional.
The lawsuit alleges that Meta’s Instagram and Google’s YouTube incorporated deliberate design tricks that manipulate users into doing what the app wants: infinite scroll that never ends, autoplay that queues the next video before users decide they want it, and notifications strategically timed to interrupt whatever else they’re doing and pull them back to the app.
Most previous cases tried to hold platforms liable for posts, images, or messages created by other users—claims that a law shielding platforms from being sued over what users post has traditionally blocked. Design choices are different. When a plaintiff alleges that Meta’s own choice to implement infinite scroll caused harm, that claim targets the platform’s own conduct, not third-party speech. This procedural distinction means the case can go forward.
The Supreme Court’s 2024 Decision
In Moody v. NetChoice, decided in 2024, the justices addressed whether states could force platforms to display all lawful speech without regard to the platform’s editorial preferences. Florida and Texas had passed laws prohibiting major social media platforms from removing content based on viewpoint. The Supreme Court struck down the laws.
This creates a problem for anyone who wants to regulate how algorithms decide what to show users. If algorithms that select and display posts constitute protected editorial expression, does a law banning the specific algorithm parameters that maximize engagement violate the Constitution by forcing platforms to change their editorial choices? The platforms will certainly argue yes.
The Commercial Speech Pathway
Speech used to sell things or promote business gets less protection than other speech. The government can regulate this kind of speech if the regulation directly advances a substantial governmental interest and solves the problem without going too far.
Addictive design mechanisms could be classified as commercial speech. The mechanisms at issue—autoplay, notification patterns, engagement metrics, recommendation algorithms—aren’t inherently expressive. They’re structural elements whose primary purpose is facilitating user engagement that drives advertising revenue. The code underlying these mechanisms carries out technical functions rather than conveying a message in the traditional sense.
If courts accepted this characterization, Congress could potentially ban or restrict addictive design mechanisms by showing that protecting child mental health is a substantial government interest, that a ban would directly advance that interest, and that the ban solves the problem without going too far. However, recent decisions have expanded constitutional protection for speaker conduct. The deeper problem is distinguishing regulable conduct from protected expression. When Instagram’s engineers design infinite scroll, they’re simultaneously making a technical engineering choice and a creative choice about how to present content to users. A regulation banning infinite scroll appears to target expression rather than conduct, pushing it into strict scrutiny territory.
The Deception Pathway
One constitutional pathway that might work involves treating addictive design mechanisms as a form of deception rather than regulating pure expression. The constitutional foundation for regulating deceptive practices is solid. Courts have long held that the government can ban false or misleading business messages.
If Congress could establish that certain design mechanisms—such as manipulating notifications to arrive when teens are most vulnerable, or arranging recommendation algorithms to promote content known to trigger anxiety in minors—constitute deception by misrepresenting what users are receiving or how algorithms function, regulation could proceed under a more forgiving standard.
The strongest version of this argument draws an analogy to slot machine regulation. States and the federal government regulate slot machine design mechanisms that exploit cognitive biases and psychological vulnerabilities. These regulations have survived constitutional challenge because they regulate conduct—the operation of gaming devices—rather than pure speech. Similarly, tobacco regulations require warnings and prohibit certain marketing techniques not because they target the truthfulness of speech, but because they prevent companies from concealing material facts through design and presentation choices.
A Congressional law could potentially prohibit design mechanisms that obscure information about how algorithms function, how much time users have spent on a platform, or what data is being collected. However, this deception-based pathway has limits. It requires establishing that the design mechanisms are misleading or deceptive, not merely engaging. A judge could conclude that while infinite scroll engages users compulsively, it doesn’t deceive them about what they’re receiving—users know they’re scrolling through content, even if the interface makes stopping difficult. Courts have said the government can’t suppress speech merely because it finds the message too persuasive.
Is Platform Design Speech or Conduct?
The threshold question is whether design mechanisms constitute “speech” at all. Courts have recognized since the 1990s that computer code itself qualifies as expressive activity protected by the Constitution. In Bernstein v. Department of Justice, a federal appeals panel held that encryption source code constitutes speech because it communicates mathematical ideas. Code is a specialized language expressing the programmer’s ideas in a format readable by computers, and this expressive dimension deserves protection even though the code also performs a technical function.
Platform algorithms fit this framework. A recommendation algorithm is code that encodes decisions about which content to display to which users, based on parameters chosen by engineers. Those parameters reflect editorial choices—decisions about what the platform considers valuable, relevant, or engagement-worthy. When an engineer tweaks the algorithm to prioritize content that keeps users on the platform longer, she’s making an expressive choice about what editorial values the algorithm will implement. Under the logic of Moody, this editorial choice receives constitutional protection even if expressed through code rather than prose.
A law banning “infinite scroll” would effectively require platforms to redesign their code to present content differently—perhaps requiring chronological feeds that don’t automatically refresh, or manually paginated content that requires user action to see more. Such a requirement appears to compel platforms to change their expression, potentially violating the constitutional rule against forcing people to say things they disagree with.
However, courts have long recognized that when an activity has both expressive and non-expressive dimensions, the government can sometimes regulate the non-expressive aspects. In United States v. O’Brien, the justices upheld conviction of a man who burned his draft card to protest the Vietnam War. The opinion held that even though the burn was intended to communicate opposition to the war, the government had a substantial interest in preserving draft cards as administrative documents, and the regulation wasn’t aimed at suppressing the message. By this logic, if Congress can establish that regulating design mechanisms serves a substantial non-speech-related interest—such as child mental health protection—and doesn’t primarily aim at suppressing the platforms’ message, the regulation might survive even if it incidentally affects expressive choices.
The Child Protection Exception
Courts have repeatedly held that the government has particularly strong interests in protecting minors, and that constitutional protection for adult speech sometimes yields to those interests. In a 1968 case, the Supreme Court upheld a law making it illegal to sell nude materials to minors, even though those same materials would be protected if sold to adults. The opinion recognized that states have “a particularly important government interest in the regulation of expression affecting the young.”
A hypothetical law drafted narrowly to protect minors could potentially survive constitutional scrutiny by emphasizing the special vulnerability of young users whose brains are still developing and whose capacity to resist manipulative design isn’t yet mature. The law could be framed not as regulating platforms’ speech generally, but as regulating platforms’ conduct toward a vulnerable population where the government’s interest is particularly compelling.
This approach borrows from successful regulation of other dangerous products marketed to minors. The government can’t ban cigarette advertising entirely, but can restrict it with particular severity when it might reach children. Similarly, the government might ban certain design mechanisms specifically on platforms where the operator knows the user is a minor, while permitting those mechanisms for adult users.
However, this doctrine has limits. The justices rejected California’s video game law even though it specifically targeted minors, finding the state hadn’t adequately proven that violent video games caused psychological harm. Congress would need substantial evidence that specific design mechanisms cause concrete psychological injury to minors. That evidence would likely come from the ongoing litigation, but even sympathetic judges might view it as contested. Regulations protecting minors still must leave open parental choice—they can’t prevent exposure to speech that parents want their children to see. A Congressional law banning design mechanisms even for users under eighteen would raise questions about whether it unfairly takes away parents’ right to decide.
How Europe Regulates Design
The European Union and United Kingdom have enacted design-focused digital regulation while operating under different rules about what counts as protected speech. The EU’s Digital Services Act, which entered enforcement in 2024, explicitly addresses design tricks that manipulate users into doing what the app wants and requires platforms to redesign their systems to prioritize user autonomy over engagement. The EU’s rule prohibits platforms from “designing, organising, or operating” their interfaces “in a way that deceives or manipulates the recipients of their service” or that “significantly interfere[s] with” users’ ability “to make autonomous and informed choices.”
The regulation targets design mechanisms that take advantage of how people naturally think: forced consent procedures with misleading button arrangements, confirmations that make cancelling services difficult, notifications designed to nag users into unwanted actions. The EU doesn’t frame these restrictions as prohibitions on speech or expression. It characterizes design tricks as commercial practices that violate consumer protection law.
The UK’s Online Safety Bill, enacted in 2023, requires platforms to undertake “risk assessments” of mechanisms affecting child safety and implement “mitigation measures” for design mechanisms that pose risks. The law doesn’t ban specific mechanisms outright, but instead imposes a legal responsibility to protect users from harm.
Both frameworks avoid constitutional complications partly because neither jurisdiction recognizes how algorithms decide what to show users as definitively protected speech in the American sense. European courts have been more willing to treat platform design as regulable commercial conduct subject to consumer protection law, without needing to determine whether design constitutes “expression” requiring strict constitutional protection.
If Congress attempted to import these regulatory models into American law, it would immediately encounter the constitutional framework that Moody established. American courts would likely classify platform design as expression—at least editorial expression—requiring strict rather than moderate judicial review. The deception angle that works in the EU consumer-protection context might fail in America if platforms argue they’re not deceiving users about anything. Users know what autoplay does. The legal responsibility to protect users from harm that the UK employs could be seen as forcing platforms to change their message.
The Vagueness Problem
Congressional legislation attempting to ban addictive design would face constitutional challenges because the law would be too unclear. What exactly would Congress ban? “Infinite scroll” might seem clear until you consider that many legitimate uses of scrolling interfaces serve expressive purposes—a photo gallery displaying artwork, a news feed showing important information, an archive presenting historical documents. Autoplay serves legitimate functions beyond engagement maximization. Platforms might use autoplay to assist disabled users or to provide uninterrupted viewing of intentional content sequences. Notification systems can serve user preferences and safety functions alongside engagement goals.
A law banning “addictive design mechanisms” without more is almost certainly so unclear that people can’t know what’s banned. Congress might try to be more specific, banning mechanisms like “infinite scroll” or “autoplay” or “periodic notifications.” But such specificity encounters a problem: the rule would catch too many things that shouldn’t be illegal. These mechanisms have legitimate, non-addictive uses, and a blanket prohibition would sweep up protected speech and conduct along with the allegedly harmful designs.
Congress can regulate tobacco advertising because the government can identify false or misleading statements and ban those statements while permitting truthful advertising. Congress can’t ban cigarettes’ appeal to smokers while protecting appeals to non-smokers. Similarly, Congress struggles to ban design mechanisms based on how engaging they are, because engagement itself isn’t the harm—the harm is addiction, and addiction is notoriously difficult to define with legal precision. Congress can’t write legislation requiring that platform designs avoid dopamine stimulation. That would be impossibly vague and would restrict massive amounts of human communication—music, attractive design, compelling storytelling, social connection.
What Congress Could Do
If Congress decides to move forward with design regulation despite constitutional concerns, several legislative approaches might offer varying levels of constitutional defensibility.
The narrowest option: expand the FTC’s power to challenge specific design patterns as deceptive or unfair practices targeting minors, without explicitly banning mechanisms but instead empowering the agency to regulate them case-by-case. This approach mirrors existing FTC authority over false advertising and allows flexibility to distinguish harmful from beneficial uses of design mechanisms. It would likely survive constitutional challenge because the FTC already possesses substantial authority to regulate deceptive commercial practices.
A slightly broader option: mandate that users choose to turn engagement-maximizing mechanisms on instead of having to turn them off. Rather than banning infinite scroll entirely, Congress could require that default settings provide the least-engaging options—chronological feeds, no autoplay, limited notifications—with users affirmatively choosing to enable more engaging mechanisms. This framing presents the law not as banning speech but as requiring truthful disclosure and preserving user autonomy. Even this approach faces constitutional arguments that mandating default settings constitutes forcing someone to say something they don’t want to say, but it’s defensible.
A more aggressive approach: ban specific design mechanisms entirely when applied to users under eighteen, exempting adult users and potentially exempting parental controls that parents explicitly enable. This careful focus on protecting the vulnerable population and accommodation for parental choice might satisfy special child-protection doctrine while addressing the narrowest legitimate government interest. However, this approach splits the user base, potentially creating technical complexities and raising questions about whether age-based restrictions on access to otherwise-legal mechanisms violate minors’ free speech rights.
The broadest option—a blanket prohibition on specified design mechanisms for all users regardless of age—would face the steepest constitutional hurdles and would be least likely to survive judicial scrutiny. If the mechanisms constitute protected editorial expression, prohibiting them triggers the toughest level of court review. If they constitute commercial speech, prohibiting them rather than merely requiring disclosure isn’t narrow enough to pass constitutional tests.
Where This Leaves Us
The Los Angeles trial exposes internal evidence that Meta and Google deliberately designed platforms to maximize teen engagement despite understanding the mental health costs. This evidence supports the moral and policy case for Congressional action to protect children from addictive design.
But the constitutional barriers to such action are substantial. They reflect deep principles about free speech in America, not mere corporate lobbying or judicial hostility to regulation. Recent decisions have established that platforms exercise genuine constitutionally protected editorial discretion when they design systems to display content in particular ways, even through algorithms. This protection extends to the design choices that allegedly addicted K.G.M. and millions of other teenagers.
Congress isn’t constitutionally precluded from regulating platform design. The pathways available are narrow. A regulatory regime focused on transparency, requiring platforms to disclose how algorithms function and giving users opt-out rights, might survive constitutional scrutiny. A regime that expanded FTC authority to challenge specific design mechanisms as deceptive or unfair practices, particularly when applied to minors, could build on existing legal frameworks and potentially avoid head-on constitutional conflicts. A narrowly tailored law applying restrictions only to users under eighteen, accommodating parental choice, and targeting demonstrable harms might satisfy child-protection doctrine.
But a broad prohibition on engaging design mechanisms for all users faces substantial constitutional obstacles that litigation outcomes alone are unlikely to remove. The trial will produce findings about platform intent and knowledge of harms. Those findings will matter politically and may influence how courts view future regulations. But the Constitution stands as a significant constraint on the regulatory authority that Congress might wish to exercise to protect children from the addictive design mechanisms that the trial record now documents.
K.G.M.’s testimony matters. The internal documents matter. The jury’s verdict will matter. But none of it automatically solves the constitutional problem that would face any lawmaker who tries to translate moral outrage into enforceable law. That problem—the tension between protecting children and protecting expression—doesn’t have an easy answer.
Even if the platforms lose in Los Angeles, they might still win in Washington.
Our articles make government information more accessible. Please consult a qualified professional for financial, legal, or health advice specific to your circumstances.