Madeline Batt is the Legal Fellow for the Tech Justice Law.

The Tech Litigation Roundup spotlights notable lawsuits and court decisions across a variety of tech-and-law issues.

Two US juries made history in March after issuing successive rulings against social media giants that had largely avoided courtroom losses. In New Mexico, a jury awarded $375 million in civil penalties against Meta for misleading the public about child sexual exploitation on its platforms in violation of state consumer protection law. The jury trial was the first brought by a state Attorney General against the social media giant. Days later, a jury in Los Angeles became the first to hold social media companies responsible for contributing to a plaintiff’s mental health harm via addictive design. The jury determined that Meta and Google-owned YouTube were liable for negligence and failure to warn related to the plaintiff's alleged social media addiction.

The cases proceeded under two different legal theories, but both verdicts provided support for an approach that scholars and advocates have proposed as a way to overcome tech companies' formidable legal shields: focusing on design, not content. Even as platform design lawsuits have already seen important courtroom victories, these first-of-their-kind jury verdicts were significant wins for the design approach.

Tech companies have long relied on two related defenses to avoid legal liability for their platforms: Section 230, which generally shields platforms from liability for third-party content posted on their sites, and the First Amendment's protections for freedom of speech. But even though social media platforms contain large amounts of protected content and speech, these lawsuits targeted harms that were based not on the content itself but on the design of the platforms containing it. The New Mexico lawsuit, for example, alleged that Instagram's algorithms proactively recommended would-be abusers' accounts to children on the app. The Los Angeles lawsuit highlighted features like infinite scroll, resulting in a pretrial ruling where Judge Carolyn B. Kuhl wrote, “the fact that a design feature like ‘infinite scroll’ impelled a user to continue to consume content that proved harmful does not mean that there can be no liability for harm arising from the design feature itself." Courts–and now juries–have agreed with advocates that these design-based harms were distinguishable from content on the apps.

K.G.M.'s trial: The design approach and product liability torts

The Los Angeles verdict came from the first bellwether trial in a California coordinated proceeding on social media addiction. Relying on product liability law, the case argued that Meta and Google negligently designed their platforms and failed to provide warnings about the platforms' risks, causing a now 19-year-old plaintiff known as K.G.M. to suffer from social media addiction and mental health harm.

The defendants repeatedly tried to get the case dismissed on Section 230 or First Amendment grounds pretrial, but plaintiffs' lawyers in the coordinated proceeding overcame those arguments by focusing on design features. At trial, the defendants, Meta and Google, sought to exploit another potential challenge of litigating platform harms in the tort law context: proving the alleged design defect or failure to warn actually caused the plaintiff's injury. They argued that other aspects of the plaintiff's life–her family relationships, her disabilities–were responsible for her mental health challenges, not social media. This strategy was not enough: the jury awarded $3 million in compensatory damages across both Defendants and another $3 million in punitive damages.

The verdict indicates the viability of product liability lawsuits against social media companies as thousands of lawsuits are pending in this coordinated proceeding, in a federal multi-district litigation scheduled for its first bellwether trial this summer, and beyond. More cases could still be filed. That puts defendants on the hook for financial liability at a scale that could meaningfully change incentives for many tech companies and result in significant changes to the company’s business practices.

The trial's outcome may also have implications for a growing number of product liability lawsuits filed against AI companies for harms caused by chatbot design (Tech Justice Law has filed several cases in various jurisdictions). Because chatbots generate text based on probability models rather than posting or curating human speech, AI companies lack the same basis that social media companies have to argue they should benefit from Section 230 and First Amendment protections. The only court so far to consider whether to apply First Amendment protections to a chatbot's outputs declined to do so, saying that chatbot outputs are not speech (Tech Justice Law represented the plaintiff in this case). AI companies can still argue that other factors in a specific plaintiff's life caused mental injuries, but Meta and Google's loss in K.G.M.'s case suggests that such arguments may not always prevail in a jury trial.

New Mexico v. Meta: The design approach and attorney general enforcement

In New Mexico, the verdict concerned violations of state consumer protection law; the jury agreed with Attorney General (AG) Raúl Torrez that Meta unlawfully deceived consumers by claiming that its platforms were safe for children when, according to evidence in the case, it was making design choices that prioritized profit at the expense of child safety. For these claims, the attorney general’s case did not depend on showing that Meta's platform design harmed any one child, but rather that Meta's statements misled the public about a broader pattern of harm.

While the precise laws at issue are New Mexico-specific, all 50 states have consumer protection statutes prohibiting deceptive practices. There are currently dozens of lawsuits by state attorneys general pending against tech companies under these laws, so the scope of liability could be substantial.

In addition to financial penalties, as public enforcers, state attorneys general often have greater authority to seek changes in companies' practices as court-ordered remedies. In a second trial in the New Mexico case, to be heard by a judge only, Attorney General Torrez has indicated he plans to seek injunctive relief to force changes to the platform.

This next phase of the litigation will provide insight into the types of remedies courts may order and, if injunctive relief is granted, which remedies prove both enforceable and effective at improving online safety. It may also surface differences among tech accountability advocates, as Attorney General Torrez indicated he will seek to restrict encrypted messaging features that many privacy and human rights advocates prize for protecting speech. While New Mexico v. Meta is the first lawsuit of its kind to reach this stage, past litigation offers insight into potential remedies.

That second trial will apply the design approach to another untested-at-trial legal theory for social media harms: public nuisance. Public nuisance claims, also pending in other cases brought by state attorneys general and municipalities, generally argue that social media companies' design features create an unreasonable interference with a collectively held right, such as the right to public health. In the New Mexico case, the attorney general alleged that Meta’s platforms “significantly interfere[] with public health, safety, peace, comfort, and convenience.” With this trial scheduled to begin May 4, the case will test how the design approach applies under a third theory of liability.

Other tech litigation developments:

• Anthropic wins preliminary injunction against Department of Defense: Tensions between the Defense Department and Anthropic over the company’s refusal to allow use of its technology for lethal autonomous weapons and domestic surveillance led to a courtroom battle over the agency designating its former contractor a supply chain risk. The case attracted significant civil society attention, with several amicus briefs filed (Tech Justice Law submitted a brief in support of neither party). The judge granted a preliminary injunction preventing the designation from taking effect; the decision is under appeal.
• Chatbot litigation expands in scope: The growing field of chatbot litigation saw multiple firsts this month. The first lawsuit alleging harms caused by Google’s chatbot Gemini was filed. A family’s suit against OpenAI marks the first chatbot lawsuit by victims of a mass casualty event, a school shooting in Canada, where the suspect reportedly exchanged violent messages with ChatGPT. And finally, an insurance company is bringing novel claims against OpenAI after facing an allegedly frivolous suit by a woman who sought advice from ChatGPT. The company has raised claims of unauthorized practice of law, abuse of process, and tortious interference with contract.
• Meta sued over smart glasses: A class action lawsuit alleges that Meta engaged in false and misleading advertising by claiming its AI-enabled glasses are “designed for privacy” when, according to the complaint, investigative journalists found that human workers in Kenya review highly intimate footage from the glasses to train Meta’s AI models.
• Legal consequences for xAI “undressing”: A Dutch court ordered X and its AI chatbot, Grok, to stop generating and distributing AI-generated, nonconsensual sexualized imagery and child pornography. The company faces daily penalties of €100,000 per Defendant for noncompliance. New lawsuits are continuing to be filed in the US, including an action by the city of Baltimore and a class action lawsuit against xAI for licensing its LLM for use in creating AI-generated child pornography.
• Content moderation researchers allege unlawful retaliation by the US: The Coalition for Independent Technology Research sued Marco Rubio and other federal officials, arguing that the Trump Administration is violating disinformation and other platform-focused researchers’ First Amendment and Fifth Amendment rights, as well as the Administrative Procedure Act, by targeting them for deportation and exclusion from the United States.
• Grammarly sued for AI “expert review” feature: A class action lawsuit alleges that Grammarly’s “expert review” feature, which presents AI-generated writing feedback as though it came from professional writers without the writers’ consent, violates the plaintiffs’ right of publicity.
• US tech regulation legal battles continue: Industry Association NetChoice filed a lawsuit to block Chicago’s tax on social media companies. Meanwhile, xAI lost its bid for a preliminary injunction against a California law requiring companies to disclose information about the data used to train their AI models, after a judge concluded that its First Amendment and Takings Clause challenges were unlikely to succeed.
• Kenya court declares mobile phone numbers part of digital identity: In a landmark ruling, a Kenyan court declared that mobile phone numbers are part of individuals’ digital identity, expanding privacy rights associated with phone numbers in the country.