TLDR:
Consumer Reports conducted experiments to test the accuracy of AI chatbots in providing health and safety information. While AI has potential, it is not always accurate. Users should use AI as a starting point but verify information from reliable sources.
Key Elements:
- Consumer Reports tested AI chatbots for accuracy in health and safety information.
- AI chatbots like Google Gemini and Microsoft Copilot provided answers that were not always entirely accurate.
- AI chatbots can be useful as a starting point for information, but users should verify information from reputable sources.
Technology and privacy researchers at Consumer Reports conducted extensive experiments to find out what AI can and cannot do for your health. They tested the accuracy of AI chatbots in providing health and safety information. While AI has potential, its accuracy is still in question.
Consumer Reports tested five AI chatbots on various health and safety questions. For example, when asked about the number of carbon monoxide detectors needed, Google Gemini provided almost correct information but erred slightly by suggesting placing a CO alarm inside an attached garage, which could damage the alarm.
When asked about filtering PFAS from tap water, AI chatbots like Microsoft Copilot lacked key details. The advice may lead consumers to purchase filters that are not specifically designed to remove PFAS.
AI chatbots also failed to provide accurate information regarding the safety of children playing with water beads. While Consumer Reports strongly advises against it due to health risks, AI chatbots like Meta AI did not highlight the significant dangers of water beads.
Consumer Reports recommends using AI as a starting point for information but stresses the importance of verifying information from reliable sources. Treat AI as an assistant rather than an expert, and don’t hesitate to ask questions multiple times to compare answers for accuracy.