Former Google Safety Chief Warns AI Could Harm Children More Than Social Media

Summary

Tom Siegel, the former founder of Google’s trust and safety team, has called for a slowdown in AI development, warning that the technology may harm children more deeply than social media. He said the industry is repeating the earlier pattern of launching powerful products while assuming it can address harms without outside help, even as millions of young people use increasingly capable chatbots. Siegel’s concerns include suicide, mental distress, and “cognitive offloading,” in which dependence on AI answers may weaken critical thinking. The warning comes after a Reuters report on an internal Meta policy document that said its chatbot could engage in romantic or sexual conversations with children; the report prompted a congressional investigation, while regulation has struggled to keep pace. Siegel has joined Common Sense Media as the first executive director of its Youth AI Safety Institute, founded in May, with support from Anthropic and the OpenAI Foundation. He plans to develop standards for holding AI companies accountable and work with companies and outside organizations. He has also urged Anthropic and OpenAI to adopt stricter age verification and suggested that Google offer switches to disable AI Overviews and AI Mode for children. Google said parents can use existing controls to disable search access and argued that AI search can help young people learn. OpenAI said it uses self-reported age together with account and behavioral signals, while acknowledging that no single age-verification method is perfect and that it continues to test ways to improve protection and prevent circumvention.