Tom Siegel, former Vice President of Trust and Safety at Google, is issuing an urgent call to the tech industry: slow down the release of new AI features until protections for minors are strengthened. Siegel, now heading the Youth AI Safety Institute at Common Sense Media, warns that the industry is at risk of repeating the same mistakes made during the social media era.
Independent Oversight and Safety Standards
The new institute, supported by Anthropic and the OpenAI Foundation, aims to conduct research and independent evaluations of AI tools. Siegel likens the need for these checks to automotive safety tests, arguing that users must have a clear understanding of risks before using a product. He advocates for more robust age verification and the ability for parents to disable AI-generated responses within search engines.
Risks to Mental Health and Learning
Concerns focus on two main areas: mental health and cognitive development.
- Mental Health: Reports highlight teenagers developing emotional dependencies on AI companions. Notably, the family of 16-year-old Adam Raine has filed a lawsuit against OpenAI, alleging a link between chatbot interactions and his suicide—claims the company is addressing in court.
- Cognitive Offloading: There are fears that systematic reliance on AI for schoolwork limits critical thinking. A small study of 54 participants noted differences in content recall when using AI, though the findings remain preliminary.
The Corporate Response
Google and OpenAI maintain they have safeguards in place. Google points to existing parental controls, while OpenAI uses account data to identify underage users, though it admits no verification method is perfect. Siegel insists that the effectiveness of these measures cannot be left solely to the companies developing the technology.
