AI in Assignment Design

Using generative artificial intelligence (AI) can be both productive and limiting—it can help students to create and revise content, yet it also has the potential to undermine the process by which students create. When incorporated effectively into assignments, generative AI can be leveraged to stimulate students' ability to apply essential knowledge and develop critical thinking skills. 

As you explore the possible uses of generative AI in your course, note that establishing a general familiarity with generative AI and being mindful of accessibility and ethical concerns will be helpful. 

The following process may help you determine how to best incorporate generative AI into your course assignments.

Affirm What You Actually Want to Assess

As you decide how you might incorporate AI into your course, it’s important to revisit your current course assessment plan, most importantly your course learning outcomes—that is, the skills and knowledge you want students to learn and demonstrate by the end of your course. Once you have a clear idea of the specific skills/knowledge you want to assess, the following questions can help determine whether or not your current assignments are effective and assessing what you want them to assess:

  • Does my assignment call for the same type of thinking skills that are articulated in my class outcomes? For example, if my course learning outcome calls for students to analyze major themes in a work, is there risk of my final assignment prompting students to do more (e.g., synthesize multiple themes across multiple works) or to do less (e.g., merely identify a theme) than this outcome? If so, there may be a misalignment that can easily be addressed.
  • Does my assignment call for the same type of thinking skills that students have actually practiced in class? For example, if I am asking students to generate a research prospectus, have I given them adequate opportunity to develop—and receive feedback on—this skill in class?
  • Depending on your discipline, is there a need for an additional course outcome that honors what students now need to know about the use of generative AI in your course/field?

Explore When & How Generative AI Can Facilitate Student Learning

Once you have affirmed your learning outcomes and ensured that your assignments are properly aligned with those outcomes, think about if, when, and how it might make sense to incorporate generative AI. Is there a way to leverage generative AI to engage students in deeper learning, provide meaningful practice, or help scaffold your assignments?

Consider the usefulness of generative AI to serve as:

  • A means of instruction and deeper learning. Examples:
    • Have students analyze AI-generated texts to articulate what constitutes “good” (and not so good) responses to prompts.
    • Have students analyze AI-generated texts and engage in error analysis to develop more nuanced and discipline-specific writing skills.
    • Leverage the use of generative AI platforms to help students become more discerning. This can help students develop the critical thinking and information literacy skills required to effectively and responsibly use such platforms.
  • A means of providing meaningful practice/feedback opportunities. Examples:
    • Have students revise AI-generated texts to develop critical thinking skills.
    • Have students engage with a generative AI platform as a tutor. 
    • Facilitate students’ responsible, self-guided use of generative AI to develop select discipline specific skills (e.g., coding in computer science courses)
  • A means of scaffolding assignments. Examples:
    • Have students use generative AI to off-load repetitive tasks.
    • Have students use generative AI to conduct preliminary analysis of data sets to confirm broad takeaways and affirm that their more nuanced analysis is heading in the right direction.

Identify When Generative AI Cannot Facilitate Student Learning

It is often the case that students cannot—or should not—leverage generative AI to promote or demonstrate their own learning. To help ensure that your assignment design highlights students’ unique perspectives and underscores the importance of a (non-generative AI informed) discipline-specific process, consider how to emphasize metacognition, authentic application, thematic connection, or personal reflection.  

Even if another part of an assignment calls for the use of generative AI, the following strategies may supplement the uses of AI highlighted above and foster deep and meaningful learning:

  • Metacognition. Examples include:
    • Have students identify the successes and challenges they experienced throughout the completion of a project.
    • Have students set incremental goals throughout a project, highlighting next steps of a discipline-specific process, resources they used, and the steps about which they are enthusiastic/nervous.
    • Have students self-assess their work, identifying strengths and weaknesses of their product/effort.
  • Authentic application. Examples include:
    • Have students engage in problem-based learning projects, ideally in authentic settings (e.g., problems that focus on our local community, real-world challenges, real-world industries, etc.).
    • Have students present projects (and engage with) authentic audiences (e.g., real stakeholders, discipline-specific research partners, native-speaking language partners, etc.)
  • Thematic/course connections. Examples include:
    • Have students connect select reading(s) to course experiences (e.g., labs, field experiences, class discussions). 
    • Leverage Canvas-based tools that promote student-to-student interactions (e.g., Hypothesis for social annotation or FeedbackFruits for peer review and feedback).
  • Personal reflection. Examples include: 
    • Have students provide a reflective rationale for choices made throughout the completion of a class project (e.g., an artist statement, response to a reflection prompt about personal relevance of source selections)
    • Have students connect course experiences/motivations to their own lived experiences.

Create Transparent Assignment Materials

Once you have thought about whether or not generative AI can be effectively incorporated into your assignments, it is important to create assignment materials that are transparent (Winkelmes, et al., 2019). Specifically, this means creating ways to communicate to students the task you are are requiring, along with its purpose and evaluative criteria:

  • Task. Students will benefit from having a clear and accessible set of directions for the project or assignment you are asking them to complete. 
  • Purpose. Students are often more motivated when they understand why a particular task is worth doing and what specific knowledge or skills they will develop by completing the assigned task.
  • Evaluative Criteria. Students benefit from having a clear sense of how their work will be evaluated and a full understanding of what good work looks like.

Communicate Your Expectations for Generative AI Use 

Regardless of the extent to which you incorporate the use of generative AI into your assignment design, it is essential to communicate your expectations to students. Sharing clear directions for assignments, communicating how students can be successful in your class, and promoting academic integrity serves both you and your students well. 

Example Assignment Policy Language for Generative AI Use

The following language on the use of generative AI may be helpful as you create directions for specific assignments. Please note that the following sample language does not reflect general, course-level perspectives on the use of generative AI tools. For sample course-level statements, see AI & Academic Integrity.

Prohibiting AI Use for a Specific Assignment

Allowing the Use of Generative AI for a Specific Assignment with Attribution

Encouraging the Use of Generative AI for a Specific Assignment with Attribution

Confer with Colleagues

There is almost always a benefit to discussing an assessment plan with colleagues, either within or beyond your department. Remember, too, that CTI offers consultations on any topic related to teaching and learning, and we are delighted to collaboratively review your course assessment plan. Visit our Consultations page to learn more, or contact us to set up a consultation.


2023 EducaUse Horizon Report | Teaching and Learning Edition. (2023, May 8). EDUCAUSE Library.

Antoniak, M. (2023, June 22). Using large language models with care - AI2 blog. Medium.

Dinnar, S. M., Dede, C., Johnson, E., Straub, C. and Korjus, K. (2021), Artificial Intelligence and Technology in Teaching Negotiation. Negotiation Journal, 37: 65-82.

Jensen, T., Dede, C., Tsiwah, F., & Thompson, K. (2023, July 27). Who Does the Thinking: The Role of Generative AI in Higher Education. YouTube. International Association of Universities. Retrieved July 27, 2023.

OpenAI. (2023, February 16.). How should AI systems behave, and who should decide?

Winkelmes, M. A., Boye, A., & Tapp, S. (2019). Transparent design in higher education 

teaching and leadership: A guide to implementing the transparency framework institution-wide to improve learning and retention. Sterling, VA: Stylus Publishing.