With the emergence of large language and other generative AI models, every field of study is having to wrestle with their impact on its established disciplinary habits and practices. In humanities fields, the task is especially urgent given how rapidly these technologies have embedded themselves in the processes and systems that mediate creative activity and knowledge production. “Humanistic AI” starts from the premise that the most productive response to these technologies is to address them head on. The project will bring together specialists from multiple domains of intellectual inquiry (literature, linguistics, philosophy, sociology, computer science) to collectively identify the challenges generative models present to these domains. Participants will seek to articulate a strategic vision for how to evaluate these models on our own disciplinary terms; to potentially enlist them as collaborators in humanistic research; and to inform their development and deployment beyond academia. In a series of meetings held over two years, we will produce a set of case studies that model this vision, and thus help to lay the conceptual and methodological ground for a rethinking of humanities research and humanistic pursuits in the wake of generative models.