AI Health Answers Favor Google and Baidu's Own Sites, Study Finds

Think asking AI for health advice gives you neutral, trustworthy sources? A new study says no. Researchers tested 1,920 health questions on Google and Baidu's AI-powered search summaries across 12 countries and 4 languages. They found that both tech giants act like stores that only recommend their own branded products, not the best options on the shelf. Instead of sending people to a wide range of hospitals, medical journals, or public health agencies, the AI often pushed users toward company-owned platforms.

This matters because where you live changes what you learn. People in smaller or less localized countries got fewer references to their own domestic health sources. But something surprising happened: when people typed the same question in their country's official language instead of English, local sources appeared 3.5 to 13.5 times more often. So the language you use isn't just a translation issue — it changes the actual medical information you receive.

The study also looked at health topics with different levels of controversy and severity, including Traditional Chinese Medicine. It showed that health warnings and disclaimers are not consistent. What one culture's AI flags as dangerous, another might present as perfectly normal. That means AI health answers are shaped by culture and corporate interests as much as by medical facts.

For regular people, the takeaway is simple: AI search results are not neutral. They can influence important health decisions — which doctor you trust, what treatment you ask for, or whether you even seek care. The researchers say we need oversight that takes culture and geography seriously, not just technical checks. Until then, it's smart to double-check what AI tells you, especially for anything health-related.