Meta’s Settlement and AI’s Impact on Youth

3 hours ago 0

Earlier this year, Meta finalized a significant settlement with state attorneys general. They agreed to pay up to $17.1 billion, pending court approval, and implement major changes in how Facebook and Instagram function for young users. This settlement followed a period of increasing evidence, whistleblower disclosures, lawsuits, and family testimonies about social media’s impact on children. Key changes Meta pledged include limits on endless scrolling, nighttime use, notifications, as well as enhanced age verification and parental controls.

It represents progress but also highlights the issues that led to this point. Children have long used products designed to capture their attention, while researchers studied their effects on sleep, mental health, relationships, learning, and development. Parents were often left to manage these risks alone. Researchers gathered evidence, and families voiced their concerns. Finally, after many children were already exposed, protective measures began to catch up.

We should avoid repeating this sequence with artificial intelligence. Nvidia CEO Jensen Huang, in an interview, emphasized that AI safety is an engineering challenge. He suggested that if a product is unsafe, companies should not release it. Enhanced containment, monitoring, and technical safeguards can manage these risks.

Huang also discussed with Ezra Klein research involving over 26,000 secondary-school students in China. The study found that using generative AI improved homework performance and reduced time spent on assignments but was linked to lower exam scores within six months. Huang acknowledged concerns about children’s diminishing skills in long division and multiplication tables, suggesting that humans have always adapted when technology rendered certain skills unnecessary.

Technology has historically altered what humans need to know. But AI promises change at an unprecedented speed and scale. The outcomes for children are still uncertain. Childhood should focus on growth rather than ease or efficiency. Learning involves more than quick solutions; it includes tasks like problem-solving, frustration tolerance, recalling information, and generating original ideas.

As AI begins reshaping childhood, children and teens are already turning to AI for companionship, emotional support, and advice. A 2026 Pew Research Center survey found 12% of U.S. teens used AI chatbots for emotional support or advice. These systems offer personalized interactions, appealing to young users, especially lonely adolescents. However, AI is not inherently harmful to children. It could support children’s learning, creativity, and knowledge access. It might aid in difficult conversations or direct children to appropriate support sources.

Yet, ‘it might help’ and ‘we can fix problems later’ should not be our only child-safety approach. Scientists are increasingly voicing similar concerns. An international study involving 272 AI experts identified risks like misinformation, overreliance, and competitive pressures among developers. Other researchers are exploring how AI could affect children’s cognitive, emotional, and social development, indicating that outcomes highly depend on age, usage, system design, adult involvement, and what AI interactions replace in a child’s life.

These distinctions matter significantly. An AI tutor helping a teenager solve an equation differs from one providing the answer. Chatbots aiding in story brainstorming differ from those writing the story. Tools encouraging distressed children to talk with trusted adults differ from those becoming the confidant themselves. Products are advancing faster than our ability to study these nuances.

Before AI becomes integral to childhood, we should develop the scientific infrastructure to understand its effects while there is still time to influence them. Investing in independent research to track children over time and providing researchers access to data on AI’s impact on learning, relationships, emotional development, and behavior is critical. Developers should collaborate with child development experts before products reach young users, not after issues arise. Safety evaluations should be developmentally specific rather than treating children as smaller adults.

Children and Screens recently partnered with researchers across various fields to map evidence around generative AI and children. One conclusion was clear: the effects will depend on what we build. The decisions made now will determine the future impact on children. The question is not whether AI will be part of their upbringing; it will. The question is whether we will establish the research, safeguards, and scientific guidance needed to understand AI’s influence on children while we still have the opportunity to shape the future.

Kris Perry is the executive director of Children and Screens.