To the Editor: We read the comments by Burke and Gwillim with great interest. 1 Recent advancements in artificial intelligence (AI) have resulted in increased applications in various areas of medicine, including mentorship and education. Although AI-based mentorship can mitigate time limitations of mentors, we must ensure that these mentorship models do not propagate harmful stereotypes. It remains vital for educators to ensure equitable use and access of these tools for all learners; biases in education lead to biases in practice.

There has been growing popularity within academic medicine of AI-assisted tools, such as ChatGPT, that may enhance learner experiences by addressing specific queries and synthesizing information into digestible formats. 2 These tools may enhance mentorship by providing realtime feedback and presenting trainees with a safe space to satisfy their curiosities. Yet, in order to provide learners with accurate training, researchers and educators must ensure that these tools are equitable; current AI models frequently assert falsehoods as factual, 2 and have been shown to perpetuate racial biases by promoting inaccurate and harmful racebased medicine. 3