Generative AI is rapidly transforming industries, and healthcare is no exception. But how are hospitals and research institutions actually using this groundbreaking technology? A new study surveyed leaders at 36 Clinical and Translational Science Award (CTSA) institutions, providing a snapshot of the current generative AI landscape in healthcare. The findings reveal a diverse range of adoption stages, from initial experimentation to full integration. Senior leaders, IT staff, and researchers are at the forefront of these initiatives, with cross-functional committees playing a crucial role in decision-making. However, the involvement of nurses, patients, and community representatives is limited, raising concerns about inclusivity and equitable access. While most institutions lean towards a centralized, top-down governance approach for generative AI, there's significant variation in how these structures are implemented. Ethical considerations like patient privacy, data security, and algorithmic bias are top priorities, but formal ethical oversight and regulatory involvement varies widely. Interestingly, funding for generative AI projects is often ad-hoc, suggesting a cautious approach to investment. Institutions are exploring a mix of open and proprietary large language models (LLMs), opting for private or on-premises deployment rather than public cloud solutions due to security concerns. Common use cases include biomedical research, medical text summarization, and data abstraction, with accuracy and reproducibility being key evaluation metrics. Despite initial optimism, challenges remain. Data security, clinician trust, AI bias, and high maintenance costs are among the biggest concerns. A substantial skills gap in the workforce highlights the urgent need for more comprehensive training. Many institutions are still in the experimentation phase with generative AI, navigating technical hurdles, funding limitations, and regulatory uncertainties. The study's insights offer valuable lessons for the broader healthcare community. Building a robust and ethical generative AI infrastructure requires not just technical expertise but also inclusive governance, comprehensive training, and careful consideration of ethical and security implications. As generative AI continues to evolve, continuous evaluation and adaptation will be essential for maximizing its potential while minimizing its risks.