AI & Assessments: A Thoughtful Approach for Faculty

AI & Assessments: A Thoughtful Approach for Faculty

AI & Assessments: A Thoughtful Approach for Faculty

      The rapid expansion of AI in education has introduced both meaningful opportunities and real challenges for teaching and learning. One of the most pressing challenges instructors are navigating is assessment. As AI tools become more accessible, many are asking: How can we leverage AI as a tool to support and evaluate student learning rather than allow it to become a substitute for student thinking? This raises an important question. Are there truly “AI-proof” or AI-resistant assessments? While it may be unrealistic to design assessments or assignments that completely prevent AI use, we can design learning experiences that emphasize process, critical thinking, and authentic demonstration of understanding.

One helpful way to approach this is by intentionally designing assessments/assignments within three broad categories:
  1. AI-Assisted assignments: AI tools are used to support specific steps in the learning process such as brainstorming, outlining, or feedback while the student’s thinking, analysis, and final product remain central.
  2. AI-Driven assignments: engagement with AI is part of the learning objective itself. Students critically evaluate, refine, prompt, or analyze AI outputs as part of demonstrating mastery.
  3. Human-Centric assignments: Minimize reliance on AI tools, emphasizing lived experiences, real-time performance, applied skills, and personal reflection. They require students to engage directly and authentically with course material.
      Rather than searching for “AI-proof” assessments, instructors can thoughtfully choose the approach that best aligns with their learning outcomes and ethical AI integration practices (Oregon State University AI Council, n.d.). As AI tools become more accurate, the goal of AI-integrated assignments should not be simply to “catch the mistake.” While evaluating inaccuracies can build critical thinking, the deeper opportunity lies in developing an AI literacy and using professional judgment (ie. analyzing reasoning, adapting output to context, prioritizing decisions, and identifying limitations). Even highly accurate AI output still requires human interpretation, ethical consideration, and contextual expertise.
Below are several alternative approaches to traditional assessments that are more difficult to outsource to AI and better aligned with meaningful student learning (Center for Innovative Teaching and Learning [CITL], n.d.).

AI-Assisted Assignment Ideas

AI supports the process, but students remain the thinkers.
  1. Writing Courses: Students generate an AI-assisted outline and submit a reflection explaining what they revised, reorganized, strengthened, or challenged — and why. Emphasis is placed on rhetorical decisions, audience awareness, and voice rather than simply correcting errors.
  2. Nursing or Allied Health: Students evaluate an AI-generated summary or care plan and:
    1. Identify areas that require clarification, prioritization, or adaptation
    2. Modify the plan for a specific patient scenario (e.g., resource limitations, co-morbidities, cultural context)
    3. Justify their professional decisions
  3. Math & Science: Students use AI to generate multiple solution methods for a complex problem. They then:
    1. Compare the reasoning and efficiency of each approach
    2. Evaluate assumptions and completeness
    3. Determine which method is most effective and defend their choice
    4. Refine or improve the strongest solution
  4. Business or Marketing: Students brainstorm campaign ideas with AI but must:
    1. Select a final strategy
    2. Justify their choice using research and course frameworks
    1. Explain trade-offs and rejected alternatives
  5. Skilled Trades: Students review AI-generated troubleshooting steps and:
    1. Evaluate alignment with manufacturer standards
    2. Identify safety considerations
    3. Adapt recommendations to real-world constraints

AI-Driven Assignment Ideas

Working with AI is the learning goal.

  1. Information Technology or Computer Science: Students refine prompts to generate code, then analyze efficiency, limitations, and ethical considerations.
  2. Education or Human Services: Students critique an AI-generated lesson plan or case scenario using established course frameworks and propose revisions that better align with developmental, ethical, or contextual considerations.
  3. Math & Science: Students generate an AI-created dataset or simulation and:
    1. Analyze underlying assumptions
    2. Interpret findings
    3. Evaluate real-world applicability and limitations
  4. Business Ethics: Students analyze AI-generated policies for bias, feasibility, stakeholder impact, and unintended consequences, explaining what they would modify before implementation.
  5. Communications: Students compare an AI-written speech to one they compose themselves and evaluate tone, audience awareness, and authenticity.

Human-Centric Assignment Ideas

Designed to highlight student voice, performance, and application.

  1. Clinical Programs: Live simulations or skill checkoffs assessed in real time.
  2. Arts Programs: Studio critiques, performances, or process portfolios documenting iterative development (CITL).
  3. Math & Science: In-lab experiments, live problem-solving sessions, or handwritten derivations where students show reasoning step-by-step.
  4. Technical Programs: In-person fabrication or applied projects evaluated on technique and safety.
  5. First-Year Seminar: Reflective journals connecting course concepts to students’ lived experiences and goals.
      As AI evolves, meaningful assignments won’t ask students to compete with it, but to think critically alongside it. Higher education’s role is to help students develop discernment, responsibility, and sound judgment around AI. By aligning assignments with clear learning outcomes, setting transparent expectations for AI use, and prioritizing critical thinking, applied skills, and authentic engagement, instructors can create assessments that truly support student learning in an evolving technological landscape.

Questions to Consider

  1. What are the actual learning outcomes I want students to demonstrate in this assignment?
  2. Does this assessment prioritize product, process, or both? Should that shift?
  3. If students used AI on this task, would they still need to demonstrate critical thinking and understanding?
  4. Have I clearly communicated expectations around AI use (allowed, limited, or not permitted)?
  5. Is AI literacy itself part of the learning goal in this course—or is deeper human application the priority?
  6. Where might reflection, revision, or real-time demonstration strengthen this assessment?
  7. How could I redesign one existing assignment to move it toward AI-Assisted, AI-Driven, or Human-Centric design?
The AISD team is here to partner with you as you explore intentional assessment design in an AI-informed landscape. Feel free to reach out to an Instructional Designer for more alternative assessment ideas!

References

Center for Innovative Teaching and Learning. (n.d.). Alternatives to traditional exams and papers. Indiana University Bloomington. https://citl.indiana.edu/teaching-resources/assessing-student-learning/alternatives-traditional-exams-papers/index.html

Oregon State University AI Council. (n.d.). AI Council. https://aicouncil.oregonstate.education/


    • Related Articles

    • Why Students Turn to AI and How to Support Their Learning

      Why Students Turn to AI and How to Support Their Learning Why do my students use AI? Many instructors are asking themselves this question. AI has entered our classrooms faster than most of us could anticipate, and responses have varied: some fear it, ...

    • A T.R.E.A.T. For Your Syllabus: An AI Syllabus Policy Framework

      Why Have an AI Use Policy? Students continue expressing confusion, fear, and uncertainty over allowable uses of Artificial Intelligence (AI) in Higher Education. Syllabi represent a reliable, go-to location where faculty can outline their positions ...

    • Thinking About AI in Your Course?

      The Risk of Letting AI " Write the Paper" Artificial intelligence is rapidly becoming part of everyday teaching and learning in higher education. Since generative AI tools such as ChatGPT became widely accessible in 2022, educators have entered a new ...

    • Can AI . . . create rubrics?

      START WITH A PROMPT When drafting the prompt, include as many specifics regarding your output expectations as possible; the details associated with a creating a grading rubric for a history course are bolded in the example below. SUGGESTED MODEL: ...

    • The T.E.A.C.H. Framework for AI Literacy

      As developers, programmers, businesses, schools, and governments infuse Artificial Intelligence (AI) into more aspects of daily operations and activity, it’s essential that users have a framework for understanding the technology. T.E.A.C.H. framework ...