Ex-Google Safety Chief Warns AI Risks to Children Exceed Social Media

The architect of Google's trust and safety operations is sounding an alarm about artificial intelligence's impact on children, arguing that the technology industry is repeating—and potentially amplifying—the mistakes that led to widespread harm from social media platforms.

Tom Siegel, who founded Alphabet's trust and safety team and served as vice president until 2019, told Reuters this week that AI companies are deploying features to young users without establishing sufficient protections. His warning comes as social media firms face thousands of lawsuits and mounting regulatory pressure over allegations they damaged minors.

"We find ourselves in a really terrible situation in terms of what AI is doing to kids," Siegel said. "The idea that the industry will figure it out by itself without outside help is unrealistic."

New role at youth safety institute

Siegel announced Tuesday he has joined Common Sense Media as the first executive director of its Youth AI Safety Institute, launched in May to assess children's AI usage. The nonprofit advocacy group, founded in 2003, counts Anthropic and the OpenAI Foundation among the institute's backers.

After leaving Google, Siegel co-founded TrustLab, a startup building safety threat detection tools for companies. He stepped aside as CEO earlier this month to take the new position, citing his belief that the tech industry had "plateaued" in creating safe online environments.

Immediate steps available

Siegel outlined concrete measures AI companies could implement now. Anthropic and OpenAI should enforce stronger age verification to prevent underage access to chatbots, he said. Google could add a toggle to disable AI Overviews and AI Mode in search results, similar to existing filters for explicit content.

The risks he cited range from suicide and psychosis to "cognitive offloading"—diminished critical thinking from over-reliance on AI for answers. While such protections might affect short-term usage or profits, companies that fail to act face greater long-term exposure to litigation and regulation, Siegel warned.

Company responses

Google responded that parents can already disable a child's search access entirely, calling AI search features "an incredibly useful way for kids and teens to learn, explore and make sense of information." OpenAI said it uses "account and behavioral signals" beyond self-reported age to identify underage users and continuously improves these systems. Anthropic did not respond to requests for comment.

Why it matters

Millions of young people already use AI chatbots daily, yet the technology is advancing faster than regulatory frameworks can adapt. Reuters previously reported that internal Meta documents acknowledged its chatbots could "engage a child in conversations that are romantic or sensual," triggering congressional investigation. Siegel's move to establish independent testing standards—comparable to automotive crash testing—represents an attempt to create accountability mechanisms before harm becomes widespread. Success depends on whether AI companies will collaborate, an outcome Siegel considers uncertain but probable enough to warrant the career shift.

At Common Sense Media, Siegel plans to develop standards for holding AI firms accountable through collaboration with companies and third-party organizations. The effectiveness of this approach remains to be seen, but the stakes are clear: without intervention, the AI industry risks replicating social media's trajectory of causing harm first and addressing consequences later.

These details were first reported by Reuters.