Institutions are using AI for all key functions. Consistent with the impact of AI in other aspects of the institution (e.g., policy), teaching and learning is the function most likely to be the focus of the institution's AI use (59%), followed closely by administration (52%) (see figure 17).

Although AI technologies can be applied to teaching and learning in a number of ways, majorities of respondents highlighted academic integrity (74%), coursework (65%), assessment practices (54%), and curriculum design (54%) as the areas of teaching and learning most impacted by AI (see figure 18).

Faculty, staff, and students are all using AI. Just as adoption across functional areas varies, so too does adoption across stakeholder groups within the institution. Student adoption of AI in particular has been a focus of respondent concern and strategic focus throughout this report, and majorities of respondents reported that students' current use of AI appears to be focused in five areas: getting answers to problems (69%); proofreading/editing their work (67%); summarizing content such as lecture notes and articles (61%); brainstorming (55%); and image or audio generation (54%) (see figure 19). Interestingly, despite institutions' focus on "workforce preparation" as the most important goal for their AI strategy, only 22% of respondents reported that students are currently using AI tools to "learn discipline-specific workforce skills."

Asked to compare student use of AI with faculty use, 68% of respondents reported that students use AI "somewhat more" or "a lot more" than faculty, while only 3% reported that faculty use AI more than students (see figure 20). This perceived imbalance in stakeholder use of AI is striking, especially when paired with the earlier finding that faculty training is much more likely to be a focus of the institution's AI-related strategy than student training (63% and 41%, respectively).

Of course, survey respondents themselves are stakeholders in their institution's adoption of AI and have found their own use cases for these technologies in their work. In fact, 86% of respondents reported that they use AI for at least one work-related task. Majorities reported that they are using AI tools for summarizing content (74%), brainstorming or ideating (71%), and creating presentations or slides (51%) (see figure 21). Uses related to personnel or human-resource issues were the least commonly selected personal uses among respondents. In open-ended comments for "Other," respondents described use cases such as accessibility compliance (e.g., writing alt text), generating images, analyzing data, coding, general writing tasks, generating synthetic data, research (e.g., conducting literature reviews), and scheduling.

With so many varied use cases across the institution's many stakeholder groups, we might expect to see widespread adoption of and institutional support for commonly used AI tools. Relatively few respondents, however, reported that their institution is providing either an institution-wide license or homegrown software for a range of common AI solutions (see figure 22)—the most common institution-wide license provided by the institution, chatbots, was reported by only 37% of respondents.