As demonstrated throughout this report, the successful adoption of AI demands a staggering array of resources and capabilities—funding, staffing, leadership, strategic and operational resilience, technology and data infrastructure—and some institutions more than others have those resources and capabilities readily at their disposal. Specifically, in the remainder of this report we explore what appears to be a widening digital AI divide, both in current practices and resources and in longer-term planning and readiness to adapt to the emerging AI landscape.

Large and small institutions share similar AI-related interests. We begin by observing that respondents from larger (i.e., 10,000 students or more) and smaller institutions (i.e., fewer than 5,000 students) do not differ significantly in their own personal use of AI technologies, their motivations for adopting AI technologies, or their expectations for AI technologies in the future. In other words, differences in AI adoption between these respondents do not appear to stem from a difference in interest, need, or outlook. Rather, respondents from larger and smaller institutions appear to be using AI tools for personal purposes at roughly the same rates (see figure 26), with "summarizing content" and "creating presentations or slides" being the only two uses with a gap of more than five percentage points. Meanwhile, only 10% of respondents from smaller institutions and 6% of respondents from larger institutions reported not using any AI-powered tools at all.

Roughly similar proportions of respondents from larger and smaller institutions prioritized the same motivations for engaging in AI-related strategic planning (see figure 27), with "the rise of student use of AI in their courses" and "risks of inappropriate uses of these technologies" as the top two motivators for both groups. "Board/trustee interest" and "Alumni interest" were the bottom two motivators for both groups.

Finally, respondents from larger and smaller institutions expressed similar outlooks for AI technologies in the next two years (see figure 28), expressing optimism for these technologies in learning analytics and in improving accessibility for students with disabilities (the two most positive areas for both groups) while largely remaining optimistic about faculty and staff trust in AI and about concerns around increased workloads (at the bottom of the list).

Experiences with and resources for AI differed appreciably between smaller and larger institutions. In their strategic planning efforts, respondents from larger institutions reported including more focus areas and were more likely to include each focus area we asked about, with the exceptions of implementing/improving data governance and implementing/improving data privacy. This gap is particularly conspicuous in areas requiring robust internal resources and infrastructure: funding AI licenses; offering AI support from IT; and creating AI applications and new AI-focused technology infrastructure (see figure 29). In addition, these respondents were significantly more likely than respondents from smaller institutions to say that their institution views AI as a strategic priority and as an investment rather than a cost. Meanwhile, respondents from smaller institutions were much more likely to report having no accommodations for new AI-related costs compared to respondents from larger institutions (70% compared to 41%, respectively).

Larger institutions are outpacing smaller institutions in policy development as well. Respondents from larger institutions were significantly more likely to report having an AUP at their institution (52% compared to 40%) (see figure 30), and respondents from smaller institutions were nearly twice as likely to report that their cybersecurity and privacy policies are "not at all adequate" (44% compared to 24%) (see figure 31).

Respondents from larger institutions were nearly three times as likely to report having a "mix of hiring new staff and upskilling existing staff" as a solution for their institution's AI-related skills needs, whereas respondents from smaller institutions were more likely to report "primarily upskilling or reskilling existing faculty or staff" (see figure 32).

With more wide-ranging AI-related strategic planning, more robust AI and technology infrastructure and AI-related supports, access to more financial and staffing resources, and the support of institutional policies and guidance, larger institutions are much better positioned now to advance new AI-focused practices and capabilities. Indeed, across the board, respondents from larger institutions were significantly more likely than respondents from smaller institutions to report using AI across institutional functions and needs (see figure 33).

Taken together, these findings paint a picture of a higher education landscape that is becoming bifurcated along a line defined by AI investment and capabilities. On one side of this divide are institutions lagging in critical infrastructure and investments—planning and anticipating but not yet arriving. On the other side are institutions that have more resources to prepare students and faculty to effectively engage with AI tools. Time will tell the extent to which this early advantage among larger institutions will solidify into a longer-term competitive advantage. For now, we can mark a divide that is present and wider than it was only a year ago, and we will continue to examine and seek to address this divide in the years ahead.