AI Council Strategic Recommendations for GenAI in Teaching and Learning

The teaching and learning subcommittee of the AI Council, with input from Information Services, should prepare quarterly or semiannual briefs for the Provost and President and Executive Steering Group. These briefs should update senior leadership on, at minimum, these topics: emergent issues at UO, debates in teaching and learning, new data on faculty and student experiences and participation in teaching development, and progress toward implementation of this strategic plan. Recognizing that GenAI is evolving too rapidly to rely on ad hoc communication, this recommendation creates a formal expectation and cycle of reporting and communication to leadership related to teaching and learning.

For the next three to five years, at minimum, the University should collect systematic data on GenAI impacts on teaching and learning, including through faculty-led projects such as TEP’s Educational Research in Action faculty group.

That group (led by College of Education’s Jenefer Husman and TEP’s Ali Söken) presented compelling findings from UO students this June that pointed to the importance of practical teaching actions—for example, explaining the rationale for policy choices, building students’ self-efficacy (the less students think they can be successful in a course, the more likely they will misuse AI), and breaking a pervasive silence on GenAI that masks both deep uncertainty and some students’ desire to exploit ambiguity.

The Rising and Flourishing Together (RAFT) survey includes questions about AI that form an important baseline for the campus; UO’s data officers should develop a regular reporting cycle to ensure these findings inform teaching and learning planning. Adding AI-related questions and prompts to our inputs about faculty teaching—the Student Experience Survey, the Instructor Reflection, and peer review—is addressed later.

Work with the University Senate to ensure GenAI course policies are added to its list of required course policies. These policies should address two dimensions:

• How the faculty member will use GenAI tools, consistent with the guardrails in the AI Policy and Principled Practice for Instruction FAQ.
• Whether and how students may use GenAI in the course.

The most effective policies include rationales grounded in specific learning outcomes and often distinguish between tasks within assignments rather than defaulting to blanket restrictions.

Properly acknowledging GenAI use is not only new to students. It’s a skill many faculty and administrators are developing—one that we should commit to modeling. By adopting a disclosure format, like this one from Monash University, we normalize good practice across our university community:

I used Copilot Chat with data protection to copyedit and review the formatting of this fully drafted GenAI in Teaching and Learning document. (See Monash’s “Acknowledging the Use of AI.”).

Work with the University Senate to adopt a strong university-wide policy requiring academic programs to account for the integrity of grades. The guiding principle: grades must reflect students’ mastery of learning outcomes, and grading schemes must ensure that students—not GenAI tools—demonstrate understanding.

Units might meet this requirement in a variety of ways, including through these examples:

• Adding new major requirements such as oral exit exams or process-oriented senior portfolios or mini-defenses.
• Certifying that a meaningful portion of required course grades is based on “live” or proctored assessments.
• Meeting relevant national accreditation standards.
• Adopting discipline-specific solutions determined appropriate by the unit (e.g., the Lundquist College of Business already requires that a “significant” portion of the course grade be based on assessments that cannot be completed using AI; the Department of Computer Science has adopted a policy to “grade the student, not the AI”).

As faculty move to ensure they’re assessing students, not AI, UO needs to adopt quality software tools for more secure assessments, including urgently needed monitoring software to undergird the integrity of exams in online asynchronous courses. Indeed, while UO has discouraged use of AI detection software up to now, new developments suggest we should be prepared to assess and pilot the latest tools. We need to anticipate these technological expenses and establish a dedicated, recurring budget line for them, rather than relying on ad hoc resourcing across units.³

Moreover, UO should build the infrastructure for in-person assessment. UO needs a larger or secondary University Testing Center. The current site, expanded in 2023 to serve a range of functions, now is almost entirely dedicated to providing accommodated test proctoring for students with disabilities—AEC now manages 12 additional overflow sites as the growing number of students with accommodations and the growing number of in-person assessments converge.⁴ UO also needs infrastructure for oral exams (a way to schedule small spaces like study carrels); some units are reporting that even the cost of paper is a hardship and drives assessment decisions, pushing units toward Canvas-based assessments when they might prefer something else.

Faculty already are running up against the constraints of class time as they shift to in-person assessment. UO should experiment with adjusted in-class contact hours for some courses, especially ones that adopt live assessment practices. Teaching, say, a 5-contact hour course instead of a 4-contact hour course should be accompanied by additional support (say, GE support or training in Gradescope) so that these courses are not more burdensome for faculty members’ total workload per course.

Some units already have stood up task forces and other bodies to address GenAI; units that don’t have such a group should identify a standing committee in each academic unit as the connection point to the AI Council and to advance local work related to GenAI policy, information sharing, and curricular evolution.

Work with the Senate’s Continuous Improvement and Evaluation of Teaching committee to consider adding a new question to the Student Experience Survey to gather course-level feedback on GenAI policies and practices. This feedback would go directly to instructors and would not be included in the formal Student Experiences of Teaching Reports used in faculty review.

The committee already has added a prompt about GenAI-related course updates to the Instructor Reflection Survey, and TEP has added a prompt about GenAI to its peer review template, helping to ensure that faculty efforts to update their teaching in light of GenAI can feature positively in reviews.

Additionally, the Provost’s Office should encourage academic leaders to ensure that all faculty reviews create occasions for reflection and supportive discussion of GenAI-related course updates.

Work with Provost’s Council to elevate knowledge and expectations around accessing teaching resources, and communicate the importance of participation to all faculty.

Charge the Office of the Provost and the Office of the Registrar with preparing a workable proposal to add a teaching-related in-service day to the academic calendar. The first such day should be centered on GenAI. This is a concrete and creative way to address the faculty concern about time, with other topics potentially coming into focus in future years.

The University should dedicate sustained funding to the most effective faculty programs, especially communities of practice. TEP and key partners in UO Libraries, IS, and UO Online will continue to collaborate on programs including the school/college Information Sessions. With Office of the Provost support, we are also launching our most ambitious effort to date: the Year of AI Action. This initiative will be anchored by a fall Day of Exchange, sustained by communities of practice across the academic year, and celebrated during a spring symposium. The Year of AI Action structure is flexible enough to absorb the next steps of several recommendations here, and may set us up to move from high-alert mode to regular maintenance mode for AI programs in future years.

Once we reach a baseline of knowledge and action—all syllabi have policies that are rationalized by learning goals and connected to departmental expectations, and faculty have considered the integrity of course grades and designed assignments less vulnerable to AI misuse—we can shift our focus more decisively toward teaching events that showcase compelling practices among colleagues with similar teaching goals.

Many faculty report lacking the foundational GenAI knowledge needed to make confident, principled decisions about GenAI in their courses. This gap is not evenly distributed: some faculty are already sophisticated users while others feel reticent to engage. The University should draw on the expertise of Information Services colleagues and leading-edge academics—such as colleagues in School of Computer and Data Science—to create low-stakes opportunities to experiment with cutting-edge AI tools and experience their capabilities firsthand.

Indeed, many colleagues—and students—decline to use AI tools at all because of deeply held concerns about their environmental and ethical implications. For these colleagues, the opportunity to experiment with local AI models (like Ollama) could be a game changer. Similarly, having the competency to create or adapt an AI tool as supplementary coach or tutor built with fundamentals of learning science in mind would allow for students to experience support while mitigating some of the cognitive risks of learning with general-purpose chatbots. Future AI programming should ensure that faculty have opportunities to see the full range of AI capabilities so they can make informed choices about whether and how AI might be meaningful for their students.

UO should offer playful occasions to use a variety of AI tools and to learn from colleagues who are making use of these tools in their teaching and research.

One of the quickest ways UO can bring AI literacy and competency instruction to all undergraduates is through the co-curriculum. Already, all incoming students complete a mandatory Ducks Have Integrity module designed and updated annually by UO Libraries, TEP, Student Conduct and Community Standards, and UO Online.

Now, colleagues in Student Services and Enrollment Management have tentatively approved a new mandatory AI Literacy and Competency module to launch in IntroDUCKtion 2027. UO should use one of the Year of AI Action communities of practice to convene faculty who already teach—or aspire to teach—AI literacy and competency courses, alongside UO Libraries colleagues, to develop a shared set of learning objectives that will guide the creation of this module. This work is especially important given research that indicates AI literacy is protective of students’ critical thinking. ⁵

Once a collaboratively developed AI literacy module is in place, the UO Composition Program is well positioned—and ready—to reinforce its content through strategic integration of relevant material into WR 121. Drawing on the distinctive strengths of the composition classroom—comparatively small groups, relational pedagogy, and a supportive environment—Composition can deepen students’ critical understanding of appropriate use of AI that enhances rather than replaces learning.

Looking further, WR 122 and WR 123 offer opportunities to scaffold targeted AI literacies for research and writing in a way that strengthens students’ capacity for self-reflection and ethical decision making related to AI use.

Led by Director Michelle Stuckey, our writing colleagues can not only deepen students’ critical AI literacy but also help UO think carefully about the future of writing instruction at UO—our most widely shared learning objective across every level of the curriculum. ⁶

Core Education Council should consider how to integrate AI literacy and competency into UO’s general curriculum. The Council should draw on the expertise of UO faculty to adopt from national frameworks, or develop specific to UO definitions and essential learning outcomes. These definitions and learning goals are important, needed common ground for the UO community and should serve as the basis of any new requirement(s). Faculty could be gathered to do this work as part of the AI Year of Action with explicit touchpoints to Core Education Council.

Moreover, Core Education Council might immediately consider drawing up examples of how AI topics could gracefully fit into UO’s methods of inquiry—written communication, critical and creative thinking, and ethical reflection.

The deepest and most durable curricular change will happen at the level of academic programs, where faculty have the expertise and authority to determine what AI competency means in their discipline. The University should create the conditions for this work to happen systematically.

Consistent with Recommendation 8 (unit-level committees), each academic unit should be asked to conduct a curriculum review that considers at least these questions: whether graduates are prepared to use AI tools appropriately in their field given disciplinary values and career outcomes; whether any program learning outcomes should be updated to reflect AI's impact on disciplinary practice; and whether existing capstone, thesis, or exit requirements adequately assess students' own knowledge and skills in an AI-enabled environment.

The Provost’s Office should provide guidance, timelines, and resources to support this review.

One compelling effect of AI is a renewed appreciation for the deeply human, connection-oriented foundations of learning. The teaching and learning of relational skills is some of the most important work a residential university can do—and many leaders across fields, including AI executives, see deep mastery of these distinctly human capacities as exactly what graduates will need in an AI-powered world. In addition, as faculty seek to assess students more authentically, live reading, writing, troubleshooting, and discussion will need to play a reinvigorated role. Currently, TEP’s Relational Pedagogies CAIT is experimenting with the teaching practices that matter most for deepening relational learning, identifying practices that scale, and articulating the learning objectives associated with them. Let’s ensure we are attending to these relational skills in equal measure to AI literacy and competency.