Former Google Security Chief Warns AI May Harm Children More Than Social Media

Tom Siegel, who founded Google’s Trust and Safety division, has urged a pause in AI advancement, warning that the tech sector is repeating the social media era’s errors and risking greater harm to children. As social media firms face massive lawsuits and settlements for endangering minors, global governments are increasingly restricting youth access to online platforms.

"We're in a very bad situation," the industry repeats its mistakes

Siegel, formerly Google’s vice president of Trust and Safety, told Reuters that the current impact of AI on children is dire. He argued that expecting the industry to self-regulate without external intervention is unrealistic. His concerns echo broader anxieties among tech experts and scholars: AI poses deeper threats to youth than social media did, yet companies are racing to integrate advanced chatbot features used by millions of young people daily.

This warning is grounded in recent events. Last year, Reuters disclosed an internal Meta policy document revealing that its chatbot could engage in romantic or sexual conversations with children, sparking a congressional investigation. Despite this, regulatory frameworks have failed to keep pace with AI’s rapid evolution. This week, Siegel joined Common Sense Media, a nonprofit dedicated to evaluating media appropriateness for children since 2003. In May, the organization launched the Teen AI Safety Institute, with Siegel as its inaugural executive director. Backed by Anthropic and the OpenAI Foundation, he aims to establish accountability standards for AI firms and foster collaboration with businesses and third-party entities.

Siegel worked at Google from 2004 to 2019, departing due to his belief that the industry had stagnated in creating safe online environments. He later co-founded TrustLab, a security startup, from which he recently stepped down as CEO. He highlighted that current AI companies are rolling out new features for children without adequate safeguards. "We are facing a very, very serious risk of harm," he stated, citing dangers ranging from suicide and mental health crises to "cognitive offloading"—the erosion of critical thinking skills as users rely on AI for answers.

Calling for Age Verification, Google and OpenAI Respond

Siegel proposed that Anthropic and OpenAI implement stricter age verification to block minors from using chatbots. He suggested Google could add toggles to disable AI overviews and search modes, similar to existing filters for explicit content. While these measures might temporarily impact usage and profits, he warned that inaction could lead to more severe lawsuits and regulatory penalties.

In response, a Google spokesperson noted that parents can completely restrict their children’s access to the search engine via control features, emphasizing that "AI search functions are a very useful way for children and teenagers to learn, explore, and understand the world." An OpenAI spokesperson stated that, beyond self-reported age, the company uses "account and behavioral signals" to identify underage users. "Although no single method of age verification is perfect, we will continue to test and improve these systems, including preventing attempts to bypass safety measures."