Teaching with AI
Practical strategies, discipline-specific examples, and proven frameworks to thoughtfully integrate AI into your courses — while preserving academic rigor and integrity.
Why Integrate AI Into Your Teaching?
AI is already in your classroom — your students are using it. The question is whether you'll shape how they use it.
Meet Students Where They Are
The vast majority of university students already use AI tools. Teaching AI literacy ensures they use these tools effectively, ethically, and with critical awareness — rather than in secret.
Deepen Critical Thinking
When students evaluate, critique, and improve AI-generated outputs, they exercise higher-order thinking skills — analysis, evaluation, and synthesis — at the top of Bloom's Taxonomy.
Amplify Your Impact
Use AI to create personalized feedback, generate varied practice problems, scaffold complex assignments, and free up time you can reinvest in mentorship and deeper discussions.
Prepare Career-Ready Graduates
Employers increasingly expect AI fluency across all fields. Teaching students to collaborate with AI — and to know its limits — is now part of preparing them for professional life.
Protect Academic Integrity
Clear AI policies and AI-inclusive assessments are more effective than prohibition. When students understand responsible use expectations, violations decrease significantly.
Model Adaptability
Faculty who engage openly with emerging technology model the intellectual curiosity and adaptability that universities prize — and students notice and respect that approach.
The goal is not to use AI everywhere — it is to make deliberate, pedagogically grounded choices about when AI adds value, when it doesn't, and how to help students develop that same judgment.
The AI Integration Spectrum
A practical framework to help you decide how and how much to integrate AI into your courses.
There is no single "right" level. Different assignments, learning objectives, and student needs call for different approaches. Most effective courses use a mix across the semester.
Restrict
AI is not permitted. Used when the learning objective requires unassisted individual performance.
- In-class handwritten exams
- Timed oral presentations
- Lab practicals and skill demonstrations
- Supervised coding assessments
Guided Use
AI is allowed with specific constraints, scaffolding, and documentation requirements.
- Use AI to brainstorm, then write independently
- Compare your draft with an AI-generated version
- Submit an AI-use reflection with the assignment
- Use only approved tools from a provided list
Explorative
Students independently experiment with AI to determine its capabilities and acceptable course uses
- Students select and compare multiple AI tools for a task
- Open-ended prompt experimentation with reflective journal
- Explore AI for research discovery and source-finding
- Document what AI does well and where it falls short
Collaborative
AI is an active part of the learning process. Students work alongside AI and evaluate outputs critically.
- Peer-review AI-generated essays for accuracy
- Use AI to generate code, then debug and optimize it
- Iterative prompt engineering exercises
- AI-assisted research with source verification
Transformative
AI enables entirely new learning activities that would not be possible without it.
- Students build custom AI tutors for a topic
- Real-time multilingual collaboration projects
- AI-powered simulation and scenario modeling
- Creating AI tools for community partners
Instructional Designs for Your College
Classroom-ready approaches tailored to each AUS college. Select your college to browse relevant examples.
College of Arts & Sciences
Humanities, social sciences, natural sciences, and mathematics
Analyzing AI "Literature": Style Forensics
Students prompt an LLM to generate text in the style of a well-known author (e.g., Virginia Woolf, Cormac McCarthy), then develop hypotheses about why the AI made certain stylistic choices. They compare against published style analyses to evaluate the AI's fidelity and blind spots.
A Tale of Two Critiques
Students read a primary source, a human-written critique, and a ChatGPT-generated critique of the same text. They compare both analyses, identifying where AI missed nuance, introduced bias, or overlooked cultural context, then write a reflection on what each approach reveals.
The AI Sandwich: Research Process
Students use AI at the beginning (brainstorming research questions, generating interview directions) and end (editing, formatting) of a research project. The core work of data collection, analysis, and original interpretation must be entirely human. They reflect on where AI helped and where it fell short.
AI-Resistant Close Reading
Assignments are anchored to in-class discourse so that essays become extensions of a conversation only those students participated in. Writing produced this way carries a voice and context that AI tools cannot replicate, making AI shortcuts both obvious and pointless. Students learn to value the process of thinking through discussion.
College of Engineering
Civil, mechanical, electrical, computer science, and industrial engineering
AI-Assisted Code, Human-Led Review
Students generate code with AI, then systematically test, debug, and refine it: writing edge-case test suites, profiling complexity, and documenting every correction. Grading weights the review process and the student's ability to explain their decisions, not the initial AI output.
Debug AI-Generated Code
Instructors or students write faulty code, then use AI to generate test cases and identify errors. Students assess the AI's suggestions, document what it missed, and submit the original code, AI response, their own fixes, and a reflection on AI as a debugging tool.
The AI Sandwich: Lab Report
Students use AI to brainstorm hypotheses and structure an outline at the start, then independently collect data, run experiments, and analyze results. AI assists only with final editing and formatting. A reflection component asks where AI added value and where the human work was irreplaceable.
AI as Teammate: Capstone Project Collaboration
Based on Mollick & Mollick's AI-teammate framework, student teams assign AI a defined role in their capstone (e.g., literature surveyor, data formatter, documentation drafter) while humans own all design decisions and analysis. Teams submit an "AI collaboration log" documenting every interaction and what they accepted, rejected, or modified.
College of Architecture, Art & Design
Architecture, visual communication, design, and multimedia
AI Image Remixing & Authorship Debate
Students upload curated images to an AI remixer tool, explore how the model combines and transforms visual inputs, then critically discuss authorship, creative boundaries, and the role of human intentionality. Documents what it means to "create" when a machine mediates the process.
AI-Assisted Concept Generation → Hand Execution
Students use AI image generation to rapidly explore dozens of visual or spatial directions during the concept phase, then develop their chosen direction entirely through manual craft, models, or digital tools. AI accelerates divergent thinking without replacing making. Based on UT Austin's Advanced Design studio approach.
Formal Critique of AI-Generated Design
Students generate design solutions using AI, then conduct a structured crit using discipline vocabulary: evaluating hierarchy, rhythm, contrast, accessibility (WCAG), cultural appropriateness, and material feasibility. The exercise reveals what AI cannot evaluate about its own output.
Parametric Form-Finding with AI
Students use AI to generate hundreds of parametric variations for an architectural element, constrained by structural and environmental parameters. Selected forms are developed in Rhino/Grasshopper or similar tools. The AI expands the solution space while the student applies judgment and craft.
School of Business Administration
Finance, marketing, management, accounting, and economics
The AI Sandwich: Case Analysis
Students use AI to brainstorm case directions and generate an outline, then independently apply course frameworks (Porter's Five Forces, VRIO, BCG Matrix). AI assists only at the end for editing. A reflection documents where AI added value and where it fell short.
Centaurs & Cyborgs: Human-AI Collaboration
Based on Ethan Mollick's "jagged frontier" research, students tackle a business problem twice: once independently, once with AI. They map where AI outperformed them and where it fell short, then develop a personal framework for when to delegate to AI versus relying on human judgment.
AI as Simulator: Stakeholder Negotiation
Using Mollick & Mollick's AI-simulator framework, students negotiate with an AI playing a specific stakeholder role (investor, regulator, dissatisfied client). The AI adapts based on the student's arguments. Students submit the full transcript plus a reflection analyzing their strategy and what they would change.
AI-Augmented Venture Pitch
Student teams build a full venture pitch using AI for market sizing, competitor mapping, financial projections, and deck design, mirroring how founders actually work today. Evaluated on strategic coherence, defensibility of assumptions, and ability to respond to live Q&A from a faculty panel.
Quick-Start Guides
Practical, step-by-step guides to implement AI in your teaching right away. Each takes under 30 minutes.
Craft a clear, specific AI policy for your course. Includes 5 ready-to-adapt templates ranging from "AI not permitted" to "AI encouraged with documentation."
- 1 Choose your integration level
- 2 Select a template and customize it
- 3 Add to your syllabus and discuss day one
Redesign an existing assignment so it develops real learning whether students use AI or not. Uses our 5-point AI-resilience checklist.
- 1 Pick an assignment to redesign
- 2 Run it through the AI-resilience checklist
- 3 Apply modifications and add rubric criteria
A complete lesson plan for a 50-minute session introducing your students to responsible AI use. Includes slides, discussion prompts, and a hands-on component.
- 1 Download the slide deck and facilitator guide
- 2 Customize for your discipline and policies
- 3 Facilitate using the guided discussion prompts
Use AI to generate detailed first-pass feedback on student work, which you then review, edit, and personalize. Can reduce grading time while improving feedback quality.
- 1 Set up your rubric as an AI prompt template
- 2 Generate initial feedback, then review and adjust
- 3 Add personal comments and deliver to students
Create a shared collection of effective prompts for your discipline. Give students starting points for using AI productively while teaching prompt engineering fundamentals.
- 1 Identify 5–10 common course tasks
- 2 Craft and test prompts for each task
- 3 Share as a course resource with usage guidelines
A structured self-assessment to evaluate how AI affects each of your existing assignments and learning outcomes. Helps you prioritize which elements to redesign first.
- 1 List all major assessments and their weight
- 2 Test each one with AI, note what it can produce
- 3 Create a priority action plan for revisions
Academic Integrity in the Age of AI
Effective approaches focus on redesigning assessments — not on surveillance and detection alone.
Require Process Documentation
Ask students to submit drafts, revision histories, and AI interaction logs alongside final work. The process reveals the learning.
Add Personal & Local Context
Tie assignments to personal experience, campus events, local data, or guest speakers — information AI can't access or fabricate convincingly.
Use Oral Components
Pair written work with brief oral defenses, presentations, or in-class discussions. Students must explain and defend their own work.
Design Multi-Stage Assessments
Break major assignments into scaffolded stages (proposal → outline → draft → final). This makes AI shortcuts visible and ensures genuine engagement.
Assess Higher-Order Thinking
Focus on evaluation, synthesis, and original analysis. Ask "why" and "how" questions that require the student's own judgment and reasoning.
Make AI Part of the Assignment
Explicitly require students to use AI, critique its output, and improve upon it. When AI use is required and documented, integrity is built in.
A Note on AI Detection Tools
Current AI detection tools have documented rates of both false positives and false negatives. Non-native English speakers are disproportionately flagged. These tools can be useful as one signal among many, but should never be the sole basis for an academic integrity charge. Focus on assignment design and process documentation for more reliable assessment.
Recommended Talks
Curated presentations from leading educators on AI in higher education.
Artificial Intelligence In Education: Teaching CS50 At Harvard With AI with Harvard University — David J. Malan
Discover how Harvard University's course integrates a 24/7 virtual "rubber duck" tutor to provide individualized support for every learner.
How AI Could Save (Not Destroy) Education — Sal Khan
The Khan Academy founder's TED talk on AI's potential to provide every student with a personal tutor and every teacher with a teaching assistant.
Practical AI for Instructors — Ethan Mollick
Wharton professor Ethan Mollick shares real classroom strategies and practical frameworks for faculty navigating AI in higher education.
Resources & Recommended Reading
Curated from leading institutions and researchers in AI and higher education.
Regularly updated, research-informed perspectives on AI's impact on education and work. Mollick's practical experiments and frameworks are widely cited by faculty worldwide. Over 400,000 subscribers.
The global framework for responsible AI integration in education. Covers policy, pedagogy, ethics, and institutional recommendations. Available in 12 languages including Arabic.
A library of real course policies, assignments, and teaching strategies from Stanford faculty. Synthesized insights on assigning, limiting, and prohibiting AI use across disciplines.
A concise, accessible primer on generative AI's implications for higher education from the leading IT and education technology association. Ideal quick read for busy faculty.
Survey findings from 500 faculty members across 200 campuses on how AI is affecting teaching, research, workload, and intellectual property. Includes policy recommendations for institutions.
Over 30 interdisciplinary research projects on AI and education, video interviews with faculty, and the annual AI+Education Summit. Includes discussion guides and the CRAFT AI literacy curriculum.
Common Faculty Concerns
Straightforward, evidence-informed responses to questions we hear often.
Ready to Start?
Pick one guide, try one activity, or simply add an AI policy to your syllabus. Every small step counts.