Meta & YouTube Guilty: Social Media Addiction Verdict

After weeks of testimony in a Los Angeles Superior Court, a California jury delivered a historic verdict on March 25, 2026: Meta and YouTube are liable for harming the mental health of a young woman identified only as KGM through their addictive platform designs. The jury assigned 70% of the liability to Meta, whose Instagram platform was at the center of the case, and 30% to YouTube. The split reflected the jury's assessment that Instagram's algorithmic recommendation system played a more direct role in driving compulsive usage. Compensatory damages totaled $3 million. But the punitive damages told the real story: $2.1 million from Meta and $900,000 from YouTube — amounts meant not to compensate the plaintiff, but to punish the companies for conduct the jury deemed reckless.

The plaintiff, now 20 years old and identified in court documents only as KGM, began using Instagram and YouTube heavily as a teenager. Her attorneys presented evidence that she became psychologically dependent on the platforms, spending hours each day consuming algorithmically curated content that reinforced negative body image, social comparison, and anxiety. KGM testified that her social media use escalated from casual browsing to compulsive behavior she could not control. She described experiencing severe depression and suicidal ideation that her medical experts linked directly to her platform usage patterns. Her therapists documented a deterioration in mental health that tracked closely with increases in screen time on both platforms. The defense argued that KGM had pre-existing mental health conditions and that social media was merely one factor among many. The jury was not persuaded.

The centerpiece of the plaintiff's case was not the content itself, but the recommendation engine that served it. Attorneys argued that Instagram's algorithm was specifically designed to maximize engagement by exploiting psychological vulnerabilities — particularly in young users. Internal Meta documents introduced as evidence showed that the company's own researchers had flagged concerns about Instagram's impact on teen mental health as early as 2019. The now-infamous internal study that found "32% of teen girls said that when they felt bad about their bodies, Instagram made them feel worse" was presented to the jury. The plaintiff's expert witnesses demonstrated how the recommendation system creates feedback loops: a user who lingers on certain content receives more of it, creating an escalating cycle of engagement that the platform profits from but the user cannot easily escape. YouTube's autoplay and recommendation sidebar faced similar scrutiny, with experts testifying that the platform's design prioritizes watch time over user wellbeing.

Both Meta and YouTube argued that Section 230 of the Communications Decency Act should shield them from liability. The law, enacted in 1996, protects internet platforms from being treated as publishers of user-generated content. The jury's rejection of this defense is significant. The plaintiff's team successfully argued that the case was not about the content users posted, but about the design decisions the companies made: how the algorithm selects, ranks, and serves content to maximize engagement. The jury agreed that algorithmic recommendation is a product design choice — not a form of speech protected by Section 230. This distinction could reshape how courts treat platform liability nationwide. If recommendation algorithms are products rather than editorial decisions, they can be subject to product liability law — the same framework used for defective cars, dangerous drugs, and unsafe consumer goods.

The verdict does not ban social media or force immediate changes to how Instagram or YouTube operate. Both companies will appeal, and the legal process could take years to reach a final resolution. However, the signal is clear: courts are beginning to treat addictive design features the same way they treat other dangerous products. This could accelerate regulatory efforts already underway in Congress and state legislatures to restrict algorithmic recommendation for minors, require age verification, and mandate parental controls. For everyday users, the practical takeaway is that the "I can't stop scrolling" experience is not a personal failing — a jury of twelve people just ruled it was engineered. Whether that leads to meaningful product changes depends on whether the tech companies decide the legal risk outweighs the engagement revenue these algorithms generate.