Higher education institutions and employers alike tout the priority they place on AI skills, but rarely do they describe what “AI skills” actually mean. These vagaries have begun to undermine some of the initial enthusiasm for integrating AI into undergraduate education, highlighting a growing need to better articulate what constellation of knowledge, behaviors, and abilities make up these skills. There is a complementary need for colleges and universities to document and understand what AI skills their instructors value and whether or not they are being taught. While there are many models of AI literacy in both educational and employment settings, there are remarkably few frameworks that identify the specific skills that make that literacy applicable in either setting. To help bridge that gap, Ithaka S+R is undertaking a new, national study to better understand what specific AI skills instructors think are important for college graduates and if they are currently teaching those skills in their classes.

The project centers on a set of almost 30 specific AI skills that fall into seven broad categories, which focus on prompt quality, the evaluation of AI outputs, ethical considerations, and managing workflows, among other topics. One of the values of this framework is the actionable granularity of the specific skills, many of which are framed using observable and assessable verbs of Bloom’s Taxonomy. For example, the AI Literacy category involves distinguishing between AI and human capabilities, selecting appropriate AI tools for specific tasks, and being able to set realistic expectations for what AI can enhance and what should remain a human task.

Another important feature of the framework is its development by human resource experts and validation by employers. The framework comes from researchers at HiBob, an international human resources technology company. HiBob has already surveyed AI decision makers, including hiring managers, corporate learning and development leaders, and executives, within a multi-national sample of 1,200 companies to better understand the importance they put on this set of skills in employment decisions and whether such skills would earn a premium for job seekers.

Ithaka S+R is currently surveying instructors of undergraduate students on the same set of AI skills across multiple disciplines and institutional types. The survey asks participants to rate the importance of each skill for college graduates in general and whether or not they personally teach each skill in their courses. The results will enable us to identify areas of convergence, and likely divergence, in the skills that college faculty are teaching and the relative value of those skills in the workforce.

The results of this project can serve as a valuable resource for academic leaders and workforce policymakers who aim to align college curricula with employers’ priorities for recent college graduates. It can also provide instructors with much-needed guidance on what learning outcomes and associated activities to include in their courses that will help their students meet the stated needs of employers. AI literacy is the foundation, but it is only through the application of specific skills are students, and eventually graduates, able to put that literacy into practice.

The results of this study will be published on our website over the summer and add to a growing body of research and technical assistance on generative AI provided by Ithaka S+R.