This session addresses the challenge of student assessment in a world with an omnipresence of AI. We present a rubric for assessing student-AI collaborative work and place it in the context of a new world where students engage in classwork using AI in new ways. In addition to empowering teachers, an overarching aim is to foster students’ responsible, skillful, and ethical use of AI and in a way that contributes to deeper learning.

AI is playing a growing and significant role in the world, still seemingly more accepted in work settings than academia. Yet, in AI is playing a growing and significant role in the world, so far more-widely adopted in work settings than academia. Yet, in educational settings, students are increasingly using generative AI tools to carry out assignments, mirroring human-AI collaboration in the workplace.

Soon after the public arrival of generative AI tools such as ChatGPT, educational institutions scrambled to create policies around AI, commonly barring student use of AI altogether. We (the presenters) considered a different position: that student use of generative AI in assignments is inevitable and even necessary to prepare students for the changing work world.

As students begin using AI to complete course assignments, instructors need a way to discern the student role in the achievement of learning objectives. This includes assessing students‘ responsible use of generative AI alongside their own demonstration of foundational skills such as critical thinking, original thinking, and creative thinking. We were driven to the challenge of designing an assessment tool by two overarching concerns: (1) a wish to see schools and instructors embrace rather than fear student use of AI; and (2) a desire to counter the potential, already beginning to manifest, for over-reliance on generative AI tools at the expense of critical thinking, creative, and original thinking. This pioneering work demonstrates a feasible path to measuring important learning objectives while promoting and assessing responsible use of generative AI.

In this education session we will share the assessment tool—the Student-AI Collaboration rubric—along with awarenesses, insights, and the journey to discovering measurable assessment criteria. Our intention is that this rubric serves as a vital, public-use building block that can be refined over time and tailored to meet unique course and institutional needs; also, that it may provide a grounding for discussion of how to re-imagine learning objectives and assignments in a human-AI collaborative world. As generative AI is changing and evolving, we expect the nature and assessment of human-AI collaboration to change, too. With the help of many others, we hope this rubric will build and grow through continued refinement.

In the presentation we share key objectives in human-AI collaboration in learning, beginning with student demonstration of critical thinking regarding AI inputs, outputs, and processes. The presentation will address the challenges, conundrums, and insights gleaned from exploring this curious new world.

In developing this rubric we drew on the AAC&U published VALUE Rubrics — Critical Thinking, Creative Thinking, and Information Literacy— along with our own original thinking. We have applied and tested the Student-AI Collaboration rubric in multiple courses across institutions, evaluated results, and made revisions. We welcome and encourage new tests and application.

Key Takeaways

• A foundational rubric that educators can use right now for assessing student-AI collaborative work, and which can be tailored to a specific course or institution.

• Deeper understanding of the challenges, opportunities, and issues associated with assessment of student use of generative AI for assignments.

• A taxonomy of objectives in human-AI collaboration in learning that can be a starting point for rich discussion within universities and colleges around the world.