AI in Course Design

When used thoughtfully, AI supports course improvements, allowing instructors to focus more on student learning outcomes. The CITT Prompt Library is a curated collection of prompts designed to help instructors, staff, and students use generative AI tools more intentionally and effectively for common teaching and learning tasks. We specifically suggest exploring the prompts are organized around the educational topics below. Interested in writing your own effective prompts for teaching and learning? Explore CITT’s AI Prompts resources for guidance, strategies, and examples to help you get started.

Assessing Student Learning

Generative AI is reshaping assessment by enabling personalized learning experiences while also challenging long-held assumptions about assessment as evidence of learning. AI can generate various assessment formats, including unique questions, simulations, and problem-solving scenarios while providing tailored feedback, targeted practice, and interactive learning opportunities.

At the same time, AI has exposed the limitations of traditional assessments as reliable measures of learning, prompting educators to reconsider what knowledge, skills, and abilities students truly need in an AI-enabled world. As instructors seek stronger evidence of student learning and competency, there is growing emphasis on assessing the learning process. This may look like requiring students to reflect on how AI was used and document their process through AI chat logs, iterative drafts, or writing transparency tools with version-history. To avoid these activities being perceived as busywork, instructors must cultivate student buy-in by positioning metacognitive practices as a key part of the learning process and an opportunity for deeper, more meaningful learning.

The following framework can help evaluate the resilience of existing assessments and identify opportunities for redesign. The 3 categories are introduced and then there are a series of reflection prompts and a rubric to help you self-evaluate.

Assessment Design Categories for the AI Era

AI-Vulnerable

Description: Assessments in this category can be completed almost entirely through direct AI prompting with minimal student thinking, synthesis, or personal application. Students can copy/paste prompts and submit AI-generated responses that meet basic requirements without demonstrating authentic learning.

Sample Assessments: Generic discussion posts, standard 5-paragraph essays on common topics, basic comprehension quizzes, formulaic lab reports

AI-Aware

Description: Assessments in this category have some protective elements but retain vulnerabilities that sophisticated AI use could exploit. They require moderate redesign or coaching to ensure students engage meaningfully with AI as a tool rather than a replacement for thinking.

Sample Assessments: Essays requiring some personal examples, projects with general guidelines, case study analyses without reflection, research papers on broad topics

AI-Resilient

Description: Assessments in this category are designed so that AI serves as a legitimate tool within the learning process rather than a shortcut around it, or through a live demonstration of learning without technology access. They emphasize unique student voice, course-specific context, iterative development, or skills that AI cannot replicate.

Sample Assessments: Reflective portfolios with process documentation, in-class presentations with Q&A, field observations with analysis, iterative projects with documented decision-making, collaborative work with individual accountability

Instructor Reflection Prompts

Is this assessment task aligned with an SLO that transparently reflects skills relevant to students in this AI Era?
Do the instructions provide clear guidance on appropriate use of AI and a compelling rationale for how the parameters support learning?
Does it integrate multiple process checkpoint documentation phases (e.g., drafts with applied feedback, conferences, chat transcripts, decision-making rationale)?
Does the task require verifiable student-specific thinking (e.g., authentic personal context, course-specific content application, local observation) or actions (e.g., live performance, presentation, in-person peer review) to minimize generic AI responses?
Are students asked to critique, evaluate, or build upon AI-generated content?
How are students prompted to reflect on their thinking or progress over time?

Assessment Analysis Rubric

Criteria	AI-Vulnerable	AI-Aware	AI-Resilient
Specificity	Generic prompts that could apply to any course or context and lack clear connection to SLOs	Some course-specific elements but could be generalized	Deeply embedded in course content, local context, or personal experience and targeting course-specific SLOs
Personal Connection	No requirement for student voice or experience	Limited personal application	Central role for unique student perspective or lived experience
Process Visibility	Only final product is evaluated with no requirement for students to document their approach or reflect on their thinking	Some intermediate checkpoints exist but with limited, or surface-level, prompting for students to articulate their reasoning and evolving approach	Multiple stages with documented thinking, feedback, and revision, using prompts that challenge students to justify decisions and push them towards deeper meaning-making
AI Role in Assignment	Fails to address AI capabilities in task design or lacks AI use parameters and rationale, allowing easy bypass of intended learning	Shows some adaptation to the AI era with basic boundaries, but lacks a compelling rationale for using or restricting AI	Eliminates ambiguity in AI’s function as one tool among many; student thinking is the dominant artifact
Verification Potential	No way to verify authentic student work	Some verification possible but not built-in	Multiple verification touchpoints (oral defense, peer knowledge, in-class work)
Skill Demonstration	Focuses on information retrieval or basic application	Requires some synthesis or analysis	Emphasizes higher-order thinking, creativity, or performance skills

Course Design Basics

Generative AI can greatly improve the processes of course design, development, and evaluation. Below are some resources that provide additional details on designing a course with highlights of different ways AI can assist with the fundamental aspects of course design:

Analyze and Design - utilize AI to brainstorm and edit student learning objectives (SLOs) and ensure alignment with course goals and assessments
Develop and Implement - AI can assist with generating assessment and facilitation ideas that align to SLOs and promote varied and inclusive representation
Evaluate and Review - analyze feedback and summarize trends and themes that provide valuable insights for continuous improvement
Universal Design for Learning (UDL) guidelines - using AI to help create unique datasets, diverse perspectives in case studies, role playing scenarios, alt text in images, and more

The Learning Process

Generative AI can help improve the learning process by integrating principles from various learning theories. By applying behaviorist, cognitivist, and constructivist principles, generative AI can create personalized learning experiences that reinforce positive behaviors, enhance information processing, and encourage active learning. For example, AI can provide immediate feedback to reinforce correct answers (behaviorism), adapt content to match a student’s cognitive load (cognitivism), and create interactive, problem-solving activities that build on prior knowledge (constructivism). Generative AI can also help when implementing the following learning strategies:

Chickering and Gamson’s Seven Principles for Good Practice in Undergraduate Education – These principles promote student success through learning community development, active learning, differentiated assessment, timely feedback, and high expectations (with support).
Robert Gagne’s 9 Events of Instruction - A structured framework for designing learning experiences that guide students from gaining attention and understanding objectives to practicing skills, receiving feedback, and transferring learning to new contexts.
Bloom’s Taxonomy - Oregon State has annotated Bloom’s 6 levels of cognitive learning with examples of how generative AI capabilities extend throughout different levels of cognitive complexity in comparison to unique human skills from basic knowledge recall to higher-order thinking skills like analysis, evaluation, and creation. By aligning AI-generated content with Bloom’s Taxonomy, educators can ensure a comprehensive and effective learning experience that promotes deep understanding and critical thinking.

Student Engagement

Generative AI can be used to take student engagement to the next level by crafting level-appropriate authentic assessments that incorporate real-life problems and scenarios for students to interact with. An instructor can extend collaboration by using generative AI to facilitate role playing for individual students or enhance established groups with an AI chatbot to fill a role in the group.

Accessible Course Design

Generative AI has the potential to unlock new and exciting course activities, but it is important to consider the accessibility of AI tools before assigning students to use them. AI is continually improving and allowing for more opportunities to support accessibility efforts, but remember to evaluate any output generated by AI before using the content in your course. Here are a few ways to get started using AI for accessibility tasks listed by level of difficulty:

Beginner - NaviGator Alt-text Generator
Intermediate - Convert captioning files to transcripts by attaching them to a generative AI tool
Advanced - Transcribe and format handwritten notes

To discover more in-depth information about evaluating tools for digital accessibility, please visit our accessible course design page.