Examining instructor attitudes to AI integration in the Cottrell Community

AI tools, including Large Language Models, are at a nascent stage of integration in physics education and physics education research. While showing early promise in tasks as diverse as providing student feedback and automated analysis of qualitative data, much is yet unknown about how these tools shape learning and what best practices for their integration might look like. Instructors inevitably will play a key role in shaping integration, and hence in this study we aimed to gain an early snapshot of their experience and perspectives on AI adoption. We chose to focus specifically on the community of Cottrell awardees, including faculty in Physics, Chemistry and Astronomy departments, from the Research Corporation for Science Advancement. Because these awards are given to early career faculty for innovative education projects as well as their research, this community of around 600 people represents a valuable pool of likely early adopters. Using a survey instrument incorporating quantitative and qualitative questions distributed to Cottrell awardees, we aimed to answer: What is the current landscape of AI adoption in STEM higher education, and how does it vary across institution types? and What factors influence STEM faculty decisions to adopt AI tools into their teaching? Initial insights into these questions will be presented together with recommendations emerging from the dataset.