Scientific Misinformation in the Age of AI

DRI is monitoring the proliferation of misleading and inaccurate information attributed to our researchers and encourages the use of trusted, reputable news sources.  

Sharing scientific findings with the broader public through the media is critical for a healthy scientific ecosystem. Unfortunately, the proliferation of AI chatbots is leading to the rise and spread of scientific misinformation. DRI has noticed a sharp increase in the number of what appear to be illegitimate websites posing as media outlets and producing articles with the use of large language models. These often discuss real issues in the news and are given an air of legitimacy by including the names of experts such as DRI’s scientists, but with fabricated quotes and research. This misinformation is then further amplified when AI chatbots, which cannot distinguish between fact and fiction, repeat the fabricated stories in response to prompts.  

Identifying the misinformation in these stories can be challenging, because they often identify sources with real expertise in the topic. For example, a chatbot may correctly identify that one of DRI’s hydrologists would be an expert source on the topic of drought in the West, note that much of their work is in the Lake Tahoe region, and produce a fabricated quote that sounds plausible. The fabrication may even include specific statistics and the names of real scientific journals. For this reason, we recommend sticking to trusted, verified media outlets that eliminate the mental gymnastics of evaluating whether an article may have been produced by chatbots. When searching for information with the use of AI, ask for the response to include a link to the original source and verify its legitimacy.  

DRI will be monitoring this rapidly evolving situation and responding accordingly. Our mission has always been to provide the most accurate and timely information possible in order to answer some of science’s biggest questions.