Concerns About AI Responses in Health Information

Democracy Dies in Darkness

Recent experiences with Google search have raised concerns about how artificial intelligence provides medical advice. An AI-generated response prompted by a medical history inquiry displayed a surprising level of authority without knowing who was asking. Another AI answer addressed a personal health question directly, also without user identification, showcasing a potential gap in handling personal information responsibly.

On September 20, 2026, Leo S. Lo of The Conversation shared an experience illustrating this issue. The Conversation, an independent nonprofit news outlet, aims to deliver factual analysis from researchers and scientists.

One day, while typing the query “How much screen time is too much for teenagers?” into Google, Lo noticed an unexpected shift in response format. Google, known for offering a list of related links, now provided an AI-generated answer. The AI cited a specific number of hours but added complexity. It emphasized the importance of the quality and balance of time spent. The response suggested that too much screen time varies for each teenager, influenced by their sleep, exercise, school tasks, and mood.

This experience raises questions about the role of AI in delivering health-related information. It highlights the need for comprehensive evaluation of AI’s capabilities in giving advice while safeguarding user data and privacy.