The convergence of recent jury verdicts, a landmark settlement, and expanded European Union oversight in 2026 marks a turning point in the fight against child exploitation online. For the first time, accountability is being built into the products themselves rather than being left to piecemeal content moderation after the harm is done.

In the first quarter of 2026, jury verdicts finding social platforms liable for negligent design, and a multibillion-dollar regulatory settlement mandated changes to how products are built. Likewise, European regulators are now subjecting AI chatbots and gaming platforms to the same scrutiny long reserved for social media.

All of this has occurred without a single new law being passed on speech or content. The pattern is consistent across courts and regulatory agencies because both are converging on product design instead of content moderation as the lever for accountability.

‍Mariana Olaizola Rosenblat, Policy Advisor on Technology & Law at NYU Stern’s Center for Business and Human Rights, has spent years studying how children are put at risk across social media, gaming, and AI chatbots. She argues that the design-focused approach is a promising development in the effort to protect children from harm in digital spaces.

The accountability shift from content to design

For more than a decade, efforts to hold platforms accountable in the United States ran into a wall. Section 230 of the Communications Act of 1934 protects online platforms from civil liability for harm caused by third-party content posted on their sites — a special type of immunity that platforms and courts often interpreted broadly to preempt almost any form of accountability. Further, the First Amendment prevents the government from abridging free speech, making content-based regulation nearly impossible in the US.

Feeling shielded from potential liability, companies optimized their designs to keep users on their platforms for as long as possible. They responded to occasional public calls for better content moderation by building up (and later quietly scaling back) their Trust & Safety teams, all while investing in engagement-driving features that boosted their bottom lines. In some cases, the same features that drove engagement — such as algorithmic recommendations, unscreened private messaging, and publicly discoverable minor profiles — created the pathways predators use to find and groom children.

However, new approaches that focus on product liability and how companies design their products have resulted in significant checks on these companies' actions. In March, a Los Angeles jury found two major social media companies negligent for designing their platforms to be addictive. A few months later, one of the two companies agreed to pay up to $17 billion to settle claims from dozens of states, alongside mandated changes to notifications, algorithmic feeds, and beauty filters.

On the regulatory front, the EU recently took a parallel path, using product design to regulate several companies that provide social media or gaming platforms, and AI chatbots. For example, over the summer, a social media company was charged with breaching EU online content rules due to design features which could expose children to predators or cyberbullying. In addition, the EU extended its designation of gaming and AI platforms to make their applications adhere to the Digital Services Act, which require companies to assess and mitigate risks of illegal content, including child sex abuse material, that can be disseminated through their platforms.

NYU's Olaizola Rosenblat explains that this regulatory pivot carries enormous significance in enabling new pathways for mitigating harm to children. "Design is the upstream mechanism that drives the harms that we see," she says. "it's like attacking the root cause rather than the symptoms."

Most importantly, this approach sidesteps the two legal barriers that killed nearly every prior effort at accountability.

The model of monetizing attention

Beneath the surface, differences among a social feed, a game, and a chatbot contain the same commercial logic because all three monetize attention. "The risk isn't the platform, it's the design choices," says Olaizola Rosenblat, adding that she counts this as good news, since design can change without pulling kids off the internet entirely.

The same design logic underlies grooming risk. Child predators need two things: first, a way to find a child; and second, a private channel to reach them. Publicly discoverable minor profiles, such as recommendation systems that bring kids to the attention of adults, and frictionless private messaging supply both.

In addition, with the proliferation of AI chatbots, the threat has evolved past needing a human predator at all. "With generative AI, the system can itself initiate sexual conversations or encourage self-harm or radicalize someone," Olaizola Rosenblat states, highlighting that this risk is now surfacing inside gaming platforms as well, with the incorporation of conversational AI characters.

Why additional protections are needed

Over the last few years, social media platforms implemented safeguards to attempt to reduce harm to children, but Olaizola Rosenblat notes that independent testers had found that many of these protections failed. In fact, she says that a majority of one major platform's advertised teen safety features were substantially ineffective or had quietly stopped working.

Oversight has similar gaps. Auditors, which tend to be  selected by the companies that they are reviewing, often set their own standards, and researcher requests for platform data can take months to process. Because of the long time frames, the product in question may introduce new features, which include design choices that negate the standard that was being researched in the first place. For gaming, where interactions are often ephemeral rather than retained, external study is even harder.

Following the example of other high-stakes safety systems, Olaizola Rosenblat suggests the creation of a permanent, well-resourced independent regulator funded through fees on the companies being regulated. "Social media is at least as consequential in children's lives as broadcast and telecom, and we regulate those with a standing agency," she explains, also calling for pre-deployment review of new features, a statutory duty of care, and outcomes measured against actual improvements in teen mental health beyond just press releases.

Despite the scale of the problem, Olaizola Rosenblat say she remains cautiously optimistic. Courts are increasingly willing to look past platforms' claims of blanket immunity; and self-regulation, she notes, is now widely understood to have failed.

What is missing, however, is federal enforcement capacity, she says. Nevertheless, she remains optimistic. "We're moving in a good direction."

You can find out more on how organizations and agencies are fighting child exploitation here