Meta’s child-safety litigation is moving from pretrial maneuvering to evidence, with opening arguments scheduled for August 18, 2026, in federal court in Oakland. The immediate stakes are broader than a single damages figure: the case could test whether states can force changes to Facebook and Instagram’s design, while adding uncertainty around the company’s financial exposure and ambitious AI spending.
That does not mean Meta owes $1.4 trillion, or that its AI plans face an automatic shutdown. The figure is the maximum exposure Meta’s lawyers say the states are seeking, while lawyers for the states have told the court that roughly $200 billion is more likely. No liability finding or final remedy has been reported in the supplied evidence.
How the California case reached trial
The action was originally filed in 2023 and consolidated into a case led by 29 state attorneys general. California Attorney General Rob Bonta is co-leading the case, which alleges that Meta encouraged addictive behavior among children and teenagers through aspects of its platform design and misrepresented product safety.
A federal judge allowed the states’ claims to proceed after Meta lost a bid to dismiss them in late June 2026. An appeals court then cleared the way for trial. The jury was seated the previous week, according to CNBC, putting the dispute on the threshold of a public examination of the states’ evidence.
The legal theory matters. Rather than centering only on user-generated content, the states are challenging design features and company conduct. The claims cite the Children’s Online Privacy Protection Act and state consumer-protection laws. That approach seeks to avoid the usual focus on Section 230 of the Communications Decency Act, a federal law that has historically shielded platforms from liability tied to user content.
The penalty debate is a warning, not a verdict
The competing numbers show how differently the two sides frame the potential consequence. Meta says the states’ demand could reach $1.4 trillion and describes that exposure as vastly disproportionate. State lawyers have indicated that $200 billion is a more likely figure. Those positions are arguments in active litigation, not an independent valuation of the company’s eventual liability.
A recent New Mexico result has raised the pressure. In a related case, a first phase produced a $375 million judgment over state unfair-practices claims. The judge also ordered $567 million into an abatement fund connected to allegations involving child sexual exploitation. Meta plans to appeal that result, so it is an important development but not a final template for the California case.
The New Mexico outcome also gives both sides a reference point. State officials can point to an existing judgment, while Meta can argue that a separate state proceeding does not establish what a federal jury in California should decide. The cases involve related concerns, but the supplied reports do not establish that their claims, evidence or remedies are identical.
Why the dispute reaches into Meta’s AI strategy
247WallSt. reported that the states want Meta to delete AI models trained on data collected from children under 13, tying the litigation to a reported $145 billion AI buildout. That is a potentially significant operational issue because model deletion, retraining or restrictions could affect compute use, data practices and development schedules. The supplied evidence does not independently verify the size of Meta’s AI investment or explain which models would be covered.
For users, the practical question is whether a legal remedy would change how Facebook and Instagram are designed for younger audiences. For Meta, the trade-off could involve compliance costs, slower product experimentation and limits on data used for model development. For regulators and other platforms, the verdict could clarify how far state consumer-protection claims can reach when they focus on product architecture instead of individual posts.
What the trial still cannot answer
The allegations remain unproven. The evidence pack does not establish whether Meta’s features caused the alleged harms, whether the company made actionable misrepresentations, or how a court would calculate damages. It also does not show that the California court has ordered any AI model to be deleted.
The next reliable signal is the trial itself: opening arguments on August 18, 2026, followed by testimony, evidence and later decisions on liability and remedies. Until those steps occur, the dollar figures and AI consequences should be treated as litigation scenarios rather than settled outcomes.
The central story is not that Meta has already incurred a trillion-dollar loss. It is that a 2026 federal trial could connect child-safety claims to the design of Facebook and Instagram, the limits of platform liability defenses, and the data practices behind Meta’s AI ambitions. The case is supported by multiple reports on its timetable, parties and legal history, but the most consequential outcomes remain contingent: a jury must assess liability, and a court would still need to determine any remedy. The New Mexico judgment provides context, not a guaranteed benchmark.
Sources and methodology
- Meta’s Lawyers Say A Loss Could Cost $1.4 Trillion. Trial That Could Determine The Future Of Zuckerberg’s AI Ambitions Is Underway. - Yahoo Finance - https://finance.yahoo.com/markets/stocks/articles/meta-stock-drops-while-1-172455887.html
- Meta's Lawyers Say A Loss Could Cost $1.4 Trillion. Trial ... - https://247wallst.com/investing/2026/08/17/metas-lawyers-say-a-loss-could-cost-1-4-trillion-trial-that-could-determine-the-future-of-zuckerbergs-ai-ambitions-is-underway
- Meta faces state AG trial over child safety claims - https://www.cnbc.com/2026/08/17/meta-attorneys-general-california-federal-trial-astronomical-consequences.html
- Meta Platforms, Inc. (META) Stock Price, News, Quote & ... - https://finance.yahoo.com/quote/META
