Artificial Intelligence

Report: Google AI Overviews Provide Dangerous Health Misinformation

Is Google's AI causing a health crisis with bad advice, or are shaming tactics making misinformation worse?

Published JAN 3

Report: Google AI Overviews Provide Dangerous Health Misinformation

Story

Above: A person holds a smartphone displaying Google AI Mode in a search query. Image credit: Smith Collection/Gado/Getty Images

The Facts

A Guardian investigation found that Google's AI Overviews, which appear at the top of search results, provided inaccurate and misleading health information in some cases. Google stated that the vast majority of its summaries are accurate and useful.
guardian
Bota Sot
Google advised pancreatic cancer patients to avoid high-fat foods, which experts described as the opposite of medical recommendations. A researcher from Pancreatic Cancer UK said this could prevent patients from consuming enough calories.
The Chosun Daily
Bota Sot
AI Overviews provided false information about liver function tests, potentially making liver disease patients wrongly believe they were healthy. The summaries displayed masses of numbers with little context and no accounting for patients' nationality, sex, ethnicity or age.
guardian
in science
Google searches for women's cancer tests provided "completely incorrect" information, including listing a pap test as a test for vaginal cancer. Experts warned this could lead women to ignore real symptoms because they had a clear result at a recent cervical screening.
guardian
in science
Health organizations and charities raised concerns that people often seek information online during moments of anxiety and uncertainty. A Google spokesperson said many examples shared were "incomplete screenshots" that linked to reputable sources and recommended seeking expert advice.
guardian
Bota Sot
Google stated that the accuracy rate of AI Overviews is comparable to other search features and that the company invests significantly in quality improvements, while experts and others said that blindly relying on AI for health issues currently remains dangerous and flawed.
Markets & Mayhem on X
in science

Sources Split

LEFT

CENTER

RIGHT

ESTAB
CRIT

CENTER

PRO
ESTAB

The Spin

Narrative A

Google's AI Overviews are delivering false and dangerous medical advice that puts lives at risk. The system wrongly tells pancreatic cancer patients to avoid high-fat foods — the exact opposite of what they need to survive treatment — and provides bogus liver test information that could leave seriously ill people thinking they're healthy. While Google bears significant responsibility, this highlights a broader public health concern: the public's over-reliance on unvetted AI for medical guidance.

guardian

Markets & Mayhem on X

Narrative B

Ridiculing people for using AI for health advice won't rebuild trust — it will repeat the COVID-era mistake where shaming backfired and hardened resistance. Not to mention, evidence shows AI can competently analyze patient histories and support triage, sometimes matching or outperforming clinicians on guidelines. People aren't irrational for exploring tools that sometimes work. Criticize bad data, not the public.

Your Local Epidemiologist

ACSH

Metaculus Prediction

Limited Coverage

This story currently has limited coverage. We will continue to monitor all major outlets and update our reporting as additional information becomes available.