A new research article from CREATE Project 3, titled “Are virtual assistants trustworthy for Medicare information: an examination of accuracy and reliability”, was published in The Gerontologist, August 2024. This study was conducted at the Florida site under the supervision of Drs. Boot and Charness with the first author being Emily Langston. This study focused on testing the accuracy and reliability of four virtual assistants using the same set of Medicare questions. Of Alexa, Google Assistant, Bard, and ChatGPT, the latter two large language models (LLM) based assistants generated more trustworthy information. When the research team performed the same analysis after a month and a half, Bard and Alexa’s accuracy remained stable, whereas ChatGPT’s decreased and Google Assistant’s increased. However, both changes were nominal in nature. Please visit the journal webpage to read the full article online.