Google puts users at risk by downplaying health disclaimers under AI Overviews

2026-02-16 · Source: AI (artificial intelligence) | The Guardian · Field: Technology & Digital — Artificial Intelligence & Machine Learning, AI Ethics & Safety · Depth: Fundamental Awareness, medium

Summary

Google is facing criticism for downplaying safety warnings within its AI Overviews, particularly concerning AI-generated medical advice. The company's AI Overviews, which appear above search results, are intended to prompt users to seek professional help for sensitive topics like health. However, an investigation by The Guardian on February 16, 2026, found that Google does not display disclaimers when users are initially presented with medical advice. Warnings only appear if users click "Show more" for additional information, and even then, they are located below the extra content in a smaller, lighter font. Experts from MIT and Queen Mary University of London express concern that this design puts users at risk of harm due to potential AI hallucinations, user input errors, and the inherent speed-over-accuracy design of AI Overviews.

Key takeaway

For AI Product Managers developing health-related features, you must prioritize explicit and prominent safety disclaimers. Your design should place warnings at the initial point of information delivery, using clear, standard-sized fonts, rather than requiring users to navigate or scroll to find them. This approach mitigates risks associated with AI inaccuracies and user over-reliance, fostering responsible AI deployment in sensitive domains.

Key insights

Google's AI Overviews downplay health disclaimers, risking user harm from AI-generated medical advice.

Principles

• Disclaimers must be prominent for critical information.
• AI models can hallucinate or prioritize user satisfaction.
• User context and question framing impact AI accuracy.

In practice

• Display safety warnings at initial content presentation.
• Ensure disclaimers are clearly visible and legible.

Topics

• Google AI Overviews
• AI Health Information
• Medical Misinformation
• AI Safety Warnings
• Generative AI Risks

Best for: AI Product Manager, Product Manager, CTO, General Interest, AI Ethicist, Tech Journalist

Related on AIssential

• Google AI Overviews put people at risk of harm with misleading health advice — Google's AI Overviews deliver inaccurate health advice, posing significant risks to user well-being.
• Google scraps AI search feature that crowdsourced amateur medical advice — Crowdsourced health advice via AI was quietly removed by Google amid scrutiny over AI-generated health information.
• Real-world safety and harms from patient-facing LLMs — Real-world data on AI chatbot harms in health is limited but growing, highlighting mental health risks.
• Users of social media and AI chatbots for health information are more likely to say they are convenient than accurate — Digital health information from social media and AI chatbots is valued for convenience and understandability, not accuracy.
• Please don’t trust your chatbot for medical advice — LLMs are unreliable for medical advice, frequently generating misinformation with overconfidence and lacking clinical reasoning.
• You can no longer Google the word ‘disregard’ — Google's aggressive AI search integration can degrade user experience for simple queries, sometimes performing worse than competitors.

Open in AIssential →

Editorial summary, takeaway, and curation by AIssential. Original article published by AI (artificial intelligence) | The Guardian.