According to CryptoSlate, a randomized trial of 400 participants published on November 4 in JAMA Network Open showed that among pregnant and postpartum individuals in the U.S., the group granted access to the maternal-infant AI chatbot Rosie showed no statistically significant advantage in self-reported anxiety and depression symptoms at 12 months compared to the control group that received monthly children's books. Participants had low overall baseline symptoms, and the Rosie intervention also included human support and incentives. The results neither established definitive benefits nor demonstrated equivalence between the two approaches, and cannot be generalized to imply that all AI mental health interventions are ineffective.