Trustworthy AI for Science: Lessons and Views From Large-Scale Experimental Facilities

Artificial Intelligence (AI) is a foundational technology of our era, playing a critical role in recent scientific advancements. However, we find ourselves increasingly concerned with the ethical, reliable, and safe application of AI in scientific contexts. In this talk, I will highlight a number of challenges around these aspects, particularly in the context of large-scale experimental facilities. I will present several examples where issues of trust and integrity are crucial, discuss our strategies for navigating these issues, and explore how collaboration within the scientific community can help us overcome these challenges.