“Navigating Artificial Intelligence in Postsecondary Education: Building Capacity for the Road Ahead,” released Tuesday by the Office of Educational Technology (OET), outlines five key recommendations and emphasizes the need for a strategic, ethical and inclusive approach:
1. DEVELOP A CLEAR VISION FOR AI INTEGRATION
This vision should ensure AI aligns with institutional goals like improving student outcomes, supporting faculty development and increasing operational efficiencies. Georgia State University, for example, uses AI to track student attendance. Complete College America’s Attainment With AI playbook offers ideas for how higher education institutions can implement generative AI, according to the OET brief.
Considerations of risk and ethics should be embedded in these policies, per the brief, to ensure AI integrations meet students’ diverse needs and avoid exacerbating inequitable outcomes or perpetuating bias.
For help considering risks and transparency, the report directs readers to the National Institute of Standards and Technology AI Risk Management Framework.
2. CREATE OR EXPAND INFRASTRUCTURE TO SUPPORT AI APPLICATIONS
Building the infrastructure to support integrating AI in higher education involves a dual approach, the brief said: developing a robust digital infrastructure and fostering human capacity.
Digital infrastructure includes reliable networks, robust cybersecurity measures to protect sensitive data, and access to sufficient computing resources. This ensures AI applications can function efficiently and securely. Equally important is development of human infrastructure, including general and discipline-specific professional development opportunities for faculty and staff to integrate AI into their work.
The report acknowledges disparities in access to necessary infrastructure between institutions. It suggests collaborative solutions like consortia and resource centers to bridge these gaps. It also highlights the energy intensity of AI models and urges institutions to consider environmental impact reduction goals in their AI strategies. TheEDUCAUSE Higher Education Generative AI Readiness Assessment can be a tool for institutions to evaluate infrastructure and identify areas for improvement.
3. RIGOROUSLY ASSESS AND EVALUATE AI-DRIVEN TOOLS, SUPPORTS AND SERVICES
Institutions must conduct thorough research and evaluation studies on AI platforms before deploying, and implement continuous improvement methods to ensure their effectiveness, safety, and alignment with educational goals and student needs.
The report distinguishes between two key elements of testing and evaluation. The first (3a) focuses on ensuring AI systems operate in alignment with ethical and privacy standards. This includes rigorous evaluation methods, stakeholder feedback and ongoing monitoring to ensure the systems align to standards, civil rights, privacy obligations and educational goals.
“Implementing robust procedures for testing AI products and services is crucial to mitigate the risk of algorithmic discrimination and ensure that AI-enabled solutions are equitable and protect student privacy,” the report said. “It is recommended that scholars conduct research in partnership with historically underserved groups to promote collaboration among AI technologists and educators to develop culturally responsive AI systems.”
The second element (3b) focuses on using research to build high-quality evidence on the effectiveness of AI systems in improving education outcomes. This includes determining what works, for whom and under what conditions. The Institute of Education Sciences’ What Works Clearinghouse offers resources for evidence-based decisions in education.
4. SEEK CROSS-INDUSTRY COLLABORATION FOR TESTING EDUCATIONAL APPLICATIONS
The report emphasizes the value of partnerships between postsecondary institutions, industry leaders, nonprofit organizations and other education stakeholders. These collaborators can enable shared expertise, resources and data to support iterative testing and design of AI models tailored to diverse educational contexts.
Institutions can leverage existing partnerships or establish new ones with organizations experienced in AI deployment to co-create solutions. The report highlights SEERnet, a collaborative network for those interested in research on digital learning platforms, and for universities with established communities of practice like Arizona State University and Georgia State University.
5. ADAPT AI OFFERINGS TO ALIGN WITH WORKFORCE NEEDS
Institutions should regularly review and revise curricula to equip students with the skills needed for an AI-driven economy, the report said. This involves integrating AI literacy into general education requirements and creating specialized programs focused on AI development, application and ethics. The University of Florida’s AI Across the Curriculum program is an example of embedding general and tailored AI education in higher ed; the report lists many other examples of colleges and universities incorporating AI into education in pharmaceuticals, philosophy and economics.
To ensure workforce readiness, it encourages working with industry partners to identify emerging skill gaps and tailor academic programs.
