Welcome! We gather at a fascinating and challenging time in education. The theme of this conference, 'Community as Resistance,' deeply resonates with our topic today. As AI tools become more prevalent, our role as educators shifts. We're not just imparting knowledge, but fostering human capacities that resist automation and commodification. The strength of our communities – our classrooms, our departments, our institutions – will be our greatest asset in this resistance.
The Current Landscape of Assignment Design
The rise of generative AI has created a pivotal moment in education, forcing a widespread re-evaluation of how we assess student learning. For decades, the take-home essay, the research paper, and the report have been mainstays of assessment. These are primarily "product-based" assignments, where the final artifact is the main object of evaluation.
AI directly challenges this model by making the production of polished text, code, or images trivial. In response, we are seeing a necessary and accelerated shift toward "process-based" assessment. This approach emphasizes the uniquely human journey of learning: the research process, the collaborative dialogue, the critical self-reflection, and the iterative development of an idea. The focus moves from "what did you produce?" to "how did you learn, and what can you do with that knowledge?" The **Tool Landscape** tab provides a detailed look at the software driving these changes, while the rest of this guide focuses on the pedagogical response.
Key Implications of this Pedagogical Shift
1. Redefining Academic Integrity
The focus of academic integrity is moving beyond a narrow definition of plagiarism toward a broader concept of ethical and transparent use of tools. Rather than simply banning AI, institutions are developing policies that require students to acknowledge and document their use of AI, shifting the conversation from "did you cheat?" to "how did you use this tool responsibly?" This requires a culture of honesty built on proactive pedagogy rather than reactive detection (Packback, 2025; Sabourin Laflamme & Bruneault, 2025).
Examples & Resources:
- University of Michigan: Course Policies & Syllabi Statements - Examples of tiered AI usage policies for syllabi.
- Mesa Community College: AI Syllabus Policy Guide - Guidance on creating transparent AI policies.
- Inside Higher Ed Report: AI and Threats to Academic Integrity - Discusses shifting student behaviors and institutional responses.
2. Emphasis on Higher-Order and Process Skills
When AI can produce a satisfactory first draft, the skills that become more valuable are those AI struggles with: critical evaluation of AI-generated content, creative synthesis of multiple sources, ethical reasoning, and personal reflection. Assessment is therefore shifting to value the *process*—drafts, critiques, reflection logs, and oral defenses—which demonstrates these durable human skills, rather than just the final, polished product (Cornell University, n.d.).
Examples & Resources:
- Stanford Teaching Commons: Integrating AI into Assignments - Strategies for leveraging multiple modalities and focusing on process.
- Utrecht University: Generative AI and Implications for Assessment - Explores adapting assessments to focus on critical engagement with AI output.
3. Heightened Concerns for Equity and Access
This shift carries significant equity implications. A reliance on technology can exacerbate the digital divide, disadvantaging students without reliable access to high-speed internet or modern devices. Furthermore, since many AI models are trained on biased data, their outputs can reinforce harmful stereotypes. An equitable approach requires designing assessments that do not depend on premium AI tools and explicitly teaching students to identify and critique algorithmic bias (Frontiers in Education, 2024; Shelton & Lanier, 2024).
Examples & Resources:
- BERA Blog: Digital Equity in the Age of Generative AI - Discusses how the digital divide now includes access to premium AI.
- Edutopia: Thinking About Equity and Bias in AI - A guide for educators on identifying and addressing bias in AI tools.
4. Urgent Need for Faculty Development
Educators cannot be expected to navigate this new landscape alone. There is an urgent need for intentional, institutionally supported faculty development. This training must go beyond simply demonstrating how to use a new tool; it must focus on pedagogical redesign, creating AI-informed assignments, and developing fair and effective assessment strategies for a new era of learning (Kolomitro & Schultz, 2024; Educause, 2024).
Examples & Resources:
- Educause: AI Literacy in Higher Education - Framework for institutional AI literacy development.
- Chronicle of Higher Education: How to Help Faculty Navigate AI - Practical strategies for faculty development programs.
In The News: Foundational Articles on AI & Pedagogy
The conversation around AI in education is evolving daily. Here are some of the precedent-setting articles from leading publications that have shaped the challenges, debates, and innovative responses in our field.
The Atlantic
The College Essay Is Dead
This seminal article by Stephen Marche was one of the first to declare that generative AI would fundamentally break traditional writing assignments, setting the stage for the entire pedagogical debate.
December 6, 2022
The New York Times
Don't Ban ChatGPT in Schools. Teach With It.
In a widely-circulated opinion piece, Kevin Roose argued against bans, proposing that educators should instead focus on teaching students critical AI literacy and how to use the tools effectively and ethically.
January 12, 2023
One Useful Thing
The Homework Apocalypse
Wharton professor Ethan Mollick's early and influential blog post framed AI not as a cheating tool but as a 'calculator for writing,' urging educators to adapt their assignments to this new reality.
October 20, 2022
The Chronicle of Higher Education
AI and the Future of Undergraduate Writing
John Warner, a key voice in writing pedagogy, provided an early, nuanced take, arguing that AI highlights pre-existing flaws in how writing is taught and assessed, urging a shift toward more meaningful, process-oriented work.
November 30, 2022
Inside Higher Ed
Professors Turn to Oral Exams and Group Work to Counter ChatGPT
This article documented the first wave of pedagogical resistance, showcasing how instructors were reviving traditional methods like oral defenses and emphasizing collaborative projects to create assessments AI couldn't solve.
February 23, 2023
The Atlantic
ChatGPT Is Dumber Than You Think
Ian Bogost provided a crucial counterpoint to the initial hype, arguing that LLMs are powerful mimics but lack true understanding, a concept that became central to designing assignments that test for genuine comprehension.
December 7, 2022
The New York Times
Alarmed by A.I. Chatbots, Universities Start Revamping How They Teach
This piece chronicled the early, system-level responses from universities, moving from outright bans to proactive pedagogical redesign, including more in-class work and handwritten assignments.
February 16, 2023
The Verge
AI-generated artwork wins state fair
An early controversy that moved beyond text, this incident at the Colorado State Fair ignited a global debate about creativity, authorship, and the role of tools in art-making, with major implications for arts and design education.
September 1, 2022
Science
ChatGPT is fun, but not an author
Leading scientific journals, including Science, issued early and influential policy statements forbidding the listing of ChatGPT as an author, establishing a key precedent for academic and scientific integrity.
January 26, 2023
The Evolving Tool Landscape in Education
The shift from product-based to process-based assessment is directly influenced by a rapidly evolving ecosystem of digital tools. Understanding these tools—what they do, how they are used, and their access models—is critical for designing relevant and effective pedagogy. Click the `[+]` next to each tool to explore its equity implications.
1. Generative AI (Content Creation)
| Tool | Primary Use | Examples in Practice | Pricing Model | Link | ||||
|---|---|---|---|---|---|---|---|---|
| ChatGPT | Conversational text, summarization, coding | Students use it as a "thought partner" to brainstorm ideas or get feedback on early drafts (Stanford, n.d.). | Freemium | Link | ||||
| ||||||||
| Google Gemini | Multimodal content generation, research | Instructors generate diverse case studies or practice problems for class discussions. | Freemium | Link | ||||
| ||||||||
| Claude (Anthropic) | Long-form writing, analysis, coding | Students use for detailed research analysis and complex problem-solving tasks. | Freemium | Link | ||||
| ||||||||
| Microsoft Copilot | Integrated AI across Office suite | Students use within Word, PowerPoint, and Excel for enhanced productivity and content creation. | Subscription | Link | ||||
| ||||||||
| Perplexity AI | Research and fact-checking with citations | Students use for initial research with automatic source citations and fact verification. | Freemium | Link | ||||
| ||||||||
| GitHub Copilot | Code completion and programming assistance | Computer science students use for learning programming concepts and debugging code. | Subscription (Free for students) | Link | ||||
| ||||||||
2. Image & Video Generation
| Tool | Primary Use | Examples in Practice | Pricing Model | Link | ||||
|---|---|---|---|---|---|---|---|---|
| DALL-E 3 | Image generation from text prompts | Art students create concept art and visual prototypes for design projects. | Freemium | Link | ||||
| ||||||||
| Midjourney | High-quality artistic image generation | Design students create mood boards and visual concepts for creative projects. | Subscription | Link | ||||
| ||||||||
| Runway ML | Video generation and editing | Film students create short video clips and experiment with visual effects. | Freemium | Link | ||||
| ||||||||
3. Academic & Research Tools
| Tool | Primary Use | Examples in Practice | Pricing Model | Link | ||||
|---|---|---|---|---|---|---|---|---|
| Grammarly | Writing assistance and grammar checking | Students use for proofreading and improving writing clarity across all disciplines. | Freemium | Link | ||||
| ||||||||
| Notion AI | Note-taking and knowledge management | Students organize research, create study guides, and collaborate on group projects. | Freemium | Link | ||||
| ||||||||
| Elicit | Research paper analysis and summarization | Graduate students use for literature reviews and finding relevant academic papers. | Freemium | Link | ||||
| ||||||||
4. Presentation & Communication
| Tool | Primary Use | Examples in Practice | Pricing Model | Link | ||||
|---|---|---|---|---|---|---|---|---|
| Gamma | AI-powered presentation creation | Students create professional presentations quickly for class projects and reports. | Freemium | Link | ||||
| ||||||||
| Beautiful.AI | Smart presentation design | Business students create investor pitches and marketing presentations with professional layouts. | Freemium | Link | ||||
| ||||||||
A Proven Framework for Human-Centered Learning: The Community of Inquiry (CoI)
Diagram adapted from Johns Hopkins Center for Teaching and Learning, based on Garrison, Anderson, & Archer (2000).
To effectively design AI-resistant learning, we can ground our practice in the Community of Inquiry (CoI) framework. This influential model posits that meaningful learning occurs in a community through the interaction of three core tenets. It provides a robust, human-centric structure perfectly suited for our goal of fostering skills AI cannot replicate.
Garrison, D. R., Anderson, T., & Archer, W. (2000). Critical inquiry in a text-based environment: Computer conferencing in higher education. *The Internet and Higher Education*, *2*(2-3), 87-105.
Social Presence
The ability of learners to project their personal characteristics into the community, thereby presenting themselves as 'real people.' This is the foundation of trust and collaboration.
AI-Resistant Applications:- Structured Peer Review: Require students to provide feedback on drafts, a nuanced task AI cannot replicate well.
- In-Class Debates or Role-Playing: Assess real-time communication and argumentation skills (Stanford, n.d.).
- Collaborative Brainstorming: Use tools like Miro or Figma where the history of collaboration is visible.
Cognitive Presence
The extent to which learners are able to construct and confirm meaning through sustained reflection and discourse. This is the process of critical inquiry itself.
AI-Resistant Applications:- Multi-Stage Assignments: Require submissions of a proposal, annotated bibliography, and draft to make the research process visible.
- AI-Critique Tasks: Have students analyze AI-generated output for flaws, biases, and weaknesses (Utrecht University, n.d.).
- Personal Connection: Ask students to connect course concepts to a personal experience or local event, requiring authentic reflection.
Teaching Presence
The design, facilitation, and direction of cognitive and social processes for the purpose of realizing personally meaningful and educationally worthwhile learning outcomes.
AI-Resistant Applications:- Targeted Human Feedback: Provide personalized feedback on higher-order concerns like argumentation and analysis.
- Facilitate Socratic Discussions: Use follow-up questions to probe for deeper understanding in class or online forums.
- Transparent Design: Explain the pedagogical rationale for your assignments and AI policies (University of Michigan, n.d.).
Three Community Strategies
Explore three actionable strategies for designing assignments that prioritize human interaction and critical thinking. Each strategy is inherently resistant to AI because it leverages processes and skills unique to human cognition and collaboration. Select a strategy below to see its concept and rationale.
Strategy 1: Cultivating Collaborative Inquiry
Concept: Assignments requiring genuine, interdependent collaboration, where the process of interaction and shared knowledge construction is as important as the product.
Why it's AI-Resistant: AI struggles with the nuanced, dynamic, and often messy process of real human interaction, negotiation, and shared ownership.
In Practice: Assignment Ideas
- Jigsaw Technique: Divide a complex topic into parts. Assign each part to a different "expert" group. Then, form new groups with one member from each expert group to share knowledge and complete a task, making learning interdependent.
- Structured Peer Review: Instead of just grading the final product, have students use a detailed rubric to provide structured feedback on each other's drafts. This assesses critical evaluation and communication skills.
- Team-Based Learning (TBL): Implement a formal TBL structure with individual readiness tests, group readiness tests, and application-focused group problem-solving. The emphasis is on the visible, interactive learning process.
Expanded Resources & Downloads
Guides & Frameworks
Tools for Collaboration
Inclusive & Equitable Design in Practice
To ensure our AI-resistant strategies serve all students, we must embed them within a culture of inclusive and equitable design. These practices ensure fairness, access, and universal support across our learning communities.
Best Practice 1: Universal Design for Learning (UDL)
Provide multiple means of engagement, representation, and action/expression. Offer choices in how students demonstrate learning.
Why it's a Best Practice:
UDL is crucial because AI proficiency is not universal. By providing multiple pathways to success, UDL ensures that assessment measures course learning, not a student's ability to master a specific AI tool. It fosters equity and builds a more inclusive community.
In Practice:
- •Offer Modality Choices: Allow a final project to be submitted as a paper, a podcast, a video, or a live presentation. This empowers students and reduces reliance on text-based AI.
- •Use AI for Accessibility: Teach students to use AI to reformat complex texts, supporting diverse learning needs. See CAST's work on AI and UDL for guidance.
- •Co-design Assessment Criteria: Involve students in a discussion about what constitutes "good work," increasing buy-in. A 2024 Educause report notes that this boosts engagement.
Best Practice 2: Transparency & Co-Design
Clearly communicate assignment purposes, expectations, and AI policies. Involve students in discussions about assessment design.
Why it's a Best Practice:
Ambiguity is the enemy of academic integrity. Transparent policies build trust and shift the classroom culture from one of adversarial detection to one of shared responsibility for learning.
In Practice:
- •Develop a Tiered AI Policy: Create a clear syllabus policy with different levels of permitted use. Many universities, like UCalgary, provide templates.
- •Require an "AI Usage Statement": Have students detail which tools they used, for what tasks, and what prompts they found most effective.
- •Explain the "Why": Include a "Pedagogical Rationale" section in assignments explaining what skills it builds and why they are valuable in a world with AI.
Best Practice 3: Address the Digital Divide & Tool Access
Be mindful of unequal access. Design assignments that don't require premium AI tools and advocate for institutional resources.
Why it's a Best Practice:
The "digital divide" now includes access to premium AI. Assuming all students can afford or access the best tools creates significant inequity. An equitable pedagogy ensures that success does not depend on a student's financial resources.
In Practice:
- •Design for the Baseline: Create assignments that can be completed successfully using only free, widely available AI tools.
- •Advocate for Institutional Licenses: Work with your library or IT department to acquire campus-wide licenses for tools like Microsoft Copilot or a private university chatbot. The University of Michigan's UM-GPT platform is a key example.
- •Provide In-Class Lab Time: For any task requiring a specific tool, dedicate class or lab time where students can use campus computers to ensure equal access.
Best Practice 4: Foster Critical AI Literacy
Teach students how to critically evaluate AI outputs, identify bias, and use AI ethically as a tool, not a crutch.
Why it's a Best Practice:
Using AI effectively is a new form of literacy. Simply banning tools leaves students unprepared for the modern workplace. Teaching them to be critical consumers of AI output is a durable skill that empowers them to be informed digital citizens.
In Practice:
- •Implement "AI Critique" Assignments: Ask students to use an AI to generate a response, then have them critique it for accuracy, bias, and omissions. This is a core practice at U of Wisconsin-Madison.
- •Teach Prompt Engineering: Dedicate a class session to demonstrating how to write effective prompts, moving from simple to complex queries to show how prompt quality impacts output quality.
- •Discuss Algorithmic Bias: Use resources like the Algorithmic Justice League to facilitate discussions about how training data can perpetuate stereotypes.
Best Practice 5: Center Process Over Product
Shift the focus of assessment from the final artifact to the observable steps of the learning journey.
Why it's a Best Practice:
When AI can easily generate a polished product, the product itself loses value as an indicator of student learning. The messy, human process—of research, drafting, revision, and collaboration—becomes the most authentic and valuable site of assessment.
In Practice:
- •Implement Multi-Stage Assignments: Break down a large research project into smaller, graded components: a proposal, an annotated bibliography, a detailed outline, a first draft, and a final version.
- •Conduct Oral Defenses or Conferences: Schedule brief one-on-one or small-group conferences where students must discuss their work, explain their reasoning, and answer questions. This makes their understanding (or lack thereof) visible.
- •Grade the Process Portfolio: Have students submit a portfolio alongside their final work that includes brainstorming notes, peer feedback, and a reflective essay on their learning journey.
Assignment Redesign Studio
This interactive toolkit guides you through redesigning a traditional assignment into an AI-resistant, human-centered learning experience by integrating the concepts from this guide.
Step 1: Select Your Current Assignment
Step 2: Select a Cognitive Goal (Bloom's Taxonomy)
Target a higher-order thinking skill. AI excels at lower-order tasks (Remember, Understand), so focus your redesign on higher-order skills to ensure academic rigor.
Remember
Recall facts
Understand
Explain ideas
Apply
Use info
Analyze
Connect ideas
Evaluate
Justify a stand
Create
Produce new work
Step 3: Choose a Core Redesign Strategy
Select one primary strategy. This will inform the types of new assignments available in the next step.
👥 Collaborative Inquiry
🌍 Authentic Engagement
🧠 Metacognition
Step 4: Select Your New Assignment
Please select a strategy in Step 3 to see assignment options.
In Action: University Case Studies
See how different institutions are putting AI-resistant principles into practice. These examples provide models for adapting assignments across various disciplines. Use the filters to find examples relevant to you.
No case studies match the selected filters.
Community Brainstorm Activities (10 min each)
Engage with these short, focused activities designed to spark conversation and collaborative thinking. In small groups, take 5 minutes for the task and 5 minutes to discuss your findings.
1. The "AI-Proof" Redesign
Task (5 min): Think of a standard assignment you currently use (e.g., a research essay). Individually, quickly list 3 ways you could redesign it to require more authentic engagement or collaborative inquiry.
Discussion (5 min): Share your redesigned ideas. Which changes seem most effective? What potential challenges or benefits for students do you foresee?
2. AI Policy Crafting
Task (5 min): Review the sample AI policy statements below. Individually, draft a single sentence for a course syllabus that clearly states your policy on using generative AI.
Discussion (5 min): Compare your policy sentences. What language is most effective for your context?
3. Community of Inquiry Self-Assessment
Task (5 min): Using the Community of Inquiry framework, rate your current course on a scale of 1-5 for each presence type. Consider specific evidence from your teaching.
Discussion (5 min): Share your ratings. Which presence type is strongest in your teaching? Which needs the most development?
🤝 Social Presence
Students feel comfortable sharing ideas and building on each other's contributions.
🧠 Cognitive Presence
Students engage in sustained inquiry and critical thinking throughout the learning process.
🧑🏫 Teaching Presence
Clear structure, facilitated discussions, and targeted feedback guide student learning.
4. Equity & Access Brainstorm
Task (5 min): Consider your student population. List 2-3 potential barriers students might face in accessing or effectively using AI tools. Then, brainstorm one concrete action you could take to address each barrier.
Discussion (5 min): Share your barriers and solutions. What institutional support would be most helpful?
Common Barriers to Consider:
- Limited access to high-speed internet or modern devices
- Financial constraints preventing access to premium AI tools
- Language barriers affecting AI prompt effectiveness
- Varying levels of digital literacy and comfort with technology
- Time constraints from work or family responsibilities
5. Process vs. Product Reflection
Task (5 min): Think about a recent assignment you graded. What percentage of the grade was based on the final product vs. the learning process? Brainstorm 2-3 ways you could make the process more visible and assessable.
Discussion (5 min): Share your process-focused ideas. What would be most feasible to implement in your context?
Process Documentation Ideas:
- Research logs with time stamps and source evaluation notes
- Draft submissions with tracked changes and revision rationales
- Peer feedback forms with structured reflection questions
- Brief video reflections explaining key decisions or challenges
- Collaborative workspace histories (Google Docs, Miro boards)
Resources for Continued Work
This work is ongoing. The following resources provide deeper insights and practical frameworks for developing AI-resistant and human-centered pedagogy.
CAST. (2024). Universal Design for Learning Guidelines.
Chang Wathall, J. (2024). AI-powered pedagogy: Redefining education.
McNulty, N. (2025, March). Inclusive design in educational AI. Niall McNulty.
MIT Sloan Teaching & Learning Technologies. (2025, May 14). 4 steps to design an AI-resilient learning experience. MIT Sloan School of Management.
University of Chicago, Generative AI. (n.d.). Strategies for designing AI-resistant assignments. Retrieved June 20, 2025, from https://genai.uchicago.edu/resources/faculty-and-instructors/strategies-for-designing-ai-resistant-assignments