Both traditional and generative AI-driven innovation is increasingly being used in clinical settings with the potential for transformative positive outcomes. Rapidly advancing AI evaluation methods can help ensure AI technologies are integrated in ways that are effective, safe, and equitable. In this talk, the Responsible AI Testing Team at DLA Piper will discuss red teaming generative AI applications in the healthcare context, such as consumer-facing AI chatbots or the use of LLMs to summarize patient-clinician interactions. We will describe recent evaluations where we applied our novel legal red teaming methodology to a healthcare chatbot. We will demonstrate how red teaming can be an e...