People are being put at risk of harm by false and misleading health information in Google’s artificial intelligence summaries, a Guardian investigation has found.
The company has said its AI Overviews, which use generative AI to provide snapshots of essential information about a topic or question, are “helpful” and “reliable”.But some of the summaries, which appear at the top of search results, served up inaccurate health information and put people at risk of harm.
In one case that experts described as “really risky”,Google wrongly advised people with pancreatic cancer to avoid high-fat foods. Experts said this was the exact opposite of what should be recommended, and may increase the risk of patients dying from the disease.
In another “alarming” example, the company provided bogus information about crucial liver function tests, which could leave people with serious liver disease wrongly thinking they are healthy.Google searches for answers about women’s cancer tests also provided “wholly wrong” information, which experts said could result in people dismissing genuine symptoms.
A Google spokesperson said that many of the health examples shared with them were “incomplete screenshots”, but from what they could assess they linked “to well-known, reputable sources and recommend seeking out expert advice”.
The Guardian investigation comes amid growing concern that AI data can confuse consumers who may assume that it is reliable. In November last year, a study found AI chatbots across a range of platforms gave inaccurate financial advice, while similar concerns have been raised about summaries of news stories.
Sophie Randall, director of the Patient Information Forum, which promotes evidence-based health information to patients, the public and healthcare professionals, said t
Google’s AI Overviews Face Criticism for Providing Inaccurate and Potentially Harmful Health Information
Table of Contents
Recent reports have revealed that Google’s AI Overviews, a new search feature designed to provide quick, summarized answers to user queries, are generating inaccurate and potentially dangerous information, particularly regarding health-related searches. Concerns have been raised by cancer charities and mental health organizations about misleading advice and inconsistent results.
Inaccurate Cancer Information
The Eve Appeal, a UK-based cancer charity, discovered that Google’s AI incorrectly listed a Pap test as a screening method for vaginal cancer. https://www.theguardian.com/technology/2024/may/16/google-ai-overviews-health-mental-health-misinformation Athena Lamnisos, the charity’s chief executive, emphasized that a Pap test screens for cervical cancer, not vaginal cancer, and providing this incorrect information could deter individuals from seeking timely medical attention for potentially serious symptoms.
“It isn’t a test to detect cancer, and certainly isn’t a test to detect vaginal cancer – this is completely wrong information,” Lamnisos stated.”Getting wrong information like this could potentially lead to someone not getting vaginal cancer symptoms checked as they had a clear result at a recent cervical screening.”
Further compounding the issue, the Eve Appeal found that the AI-generated summaries varied with each identical search, pulling information from different sources and delivering inconsistent responses. This lack of consistency raises concerns about the reliability of the feature.
Misleading Mental Health Advice
The inaccuracies extend to mental health searches. Mind, a UK mental health charity, found that AI Overviews offered “very dangerous advice” and “incorrect, harmful or could lead people to avoid seeking help” for conditions like psychosis and eating disorders. https://www.theguardian.com/technology/2024/may/16/google-ai-overviews-health-mental-health-misinformation
Stephen Buckley, Head of Information at Mind, highlighted the omission of crucial context and nuance in the AI summaries.He also noted the potential for the AI to direct users to inappropriate resources and perpetuate existing biases and stigmas surrounding mental health.
Google’s Response and Ongoing Improvements
Google acknowledges the concerns and maintains that the majority of its AI overviews are factual and helpful. The company states that the accuracy rate is comparable to its existing search features,such as featured snippets,which have been in use for over a decade. https://www.theguardian.com/technology/2024/may/16/google-ai-overviews-health-mental-health-misinformation
A Google spokesperson affirmed that the company “invests considerably in the quality of AI Overviews, particularly for topics like health” and will take corrective action when the AI misinterprets web content or lacks context, in accordance with its policies.
It is indeed significant to note that AI-powered search features are still under progress and refinement. Users seeking health information should always consult with qualified healthcare professionals and rely on reputable sources like the National Health Service (NHS) https://www.nhs.uk/ or the Centers for Disease Control and Prevention (CDC) https://www.cdc.gov/ for accurate and personalized guidance.