Back to news Wellbeing
16 July 2026 · 2 min read

The Debate Shifts From Age Limits to Safer Design in Children's Use of AI

A tech philanthropy argues that protecting children from artificial intelligence should focus on how products are designed, not only on age checks and access bans. With nearly nine in ten children and teenagers already using AI, the debate turns to whether companies, rather than parents, should carry the burden of safety.

As artificial intelligence becomes a daily presence in young lives, a discussion has emerged over where the responsibility for children's safety should lie. In an interview with PBS NewsHour, Michele Jawando, chief executive of the Omidyar Network, argued that the familiar questions of access and age verification are only the beginning of the conversation, and that the emphasis should move toward the design of the products themselves.

Jawando noted that nearly nine in ten children and teenagers are already using AI, and that roughly a quarter of young people use it every day for some purpose. In her view, banning access is not realistic, and the burden of restriction falls unfairly on families. She said the world's best engineers, rather than parents left to work it out alone, are better placed to build safer systems.

The proposed guardrails include restricting manipulative and addictive practices, particularly the encouragement of romantic attachment, building in what she described as brakes, and limiting the collection and sale of young people's data. Jawando also raised the tension in age verification: asking private companies to handle personal identification when there is no broad privacy regulation to protect that information. She pointed to age estimation and privacy built into applications from the outset as alternatives.

On regulation, she welcomed the House passage of the KIDS Safety Act as a starting point, while noting that much of it still relies on parental notification and omits a duty of care, the legal principle that a company should design a product as safely as it can. She added that some companies have taken steps, citing tools intended to notify parents when a system believes a teenager may be considering self-harm.

Jawando also questioned whether education itself must adapt, observing that schools and universities have not yet reckoned with how learning, teaching, and assessment should change in a world where AI is constantly at hand. The original reporting was produced by PBS NewsHour and can be found at PBS NewsHour.

Sources

Newsletter

Occasional news on the Foundation's programs, research, and events.

Unsubscribe at any time. See the privacy policy.