AI@UW awards 36 SEED-AI grants in its inaugural funding call

Project team
Ali Karimirad, Assistant Teaching Professor, Economics, College of Art and Sciences
A PhD student from the Economics department will be hired at a later stage

Project summary
This project develops an AI-Powered Personalized Feedback and Tutoring System (AI-PFTS) for Econ 201: Introduction to Macroeconomics, a high-enrollment course serving over 1,000 students annually. The system leverages course-specific materials—including lectures, videos, and grading rubrics—to deliver customized, on-demand academic support. By combining personalized quizzing with immediate feedback, the AI-PFTS identifies individual learning gaps and helps students strengthen core concepts more efficiently. The project also evaluates whether AI-driven support can replicate proven strategies for improving equity in education, such as deliberate practice and low-stakes assessment. In addition to enhancing learning outcomes, the system reduces student costs by replacing expensive publisher platforms and increases accessibility for students with scheduling constraints or participation barriers. The model is scalable and designed for broader institutional adoption.

Project team
Ian Schnee, Teaching Professor, Department of Philosophy, College of Arts and Sciences
Adrian KC Lee, Professor, Department of Speech and Hearing Sciences, College of Arts and Sciences

Project summary
We will develop an interactive AI tool to help instructors create clear, course-specific policies governing students’ AI use. Instead of relying on static examples or generic guidance, the tool will draw on a transparently curated library of UW policies, relevant laws, best practices from other universities and research on effective teaching. Faculty will describe their course goals, assessments and teaching values, and the AI will support a structured reflection process that results in a tailored draft AI-use policy for their syllabus. Importantly, instructors remain fully in control of the final policy. By lowering the time and effort required to create thoughtful, transparent AI guidelines, this project supports instructor autonomy, improves communication with students and promotes coherent, evidence-based approaches to AI across courses and disciplines. Students, who increasingly expect explicit guidance on how AI intersects with their coursework, stand to benefit directly through clearer expectations and more informed enrollment decisions.

Project team
Duong T. Than, Associate Teaching Professor, Sciences and Mathematics, School of Interdisciplinary Arts & Sciences, UW Tacoma

Project summary
Introductory mathematics courses often act as critical gatekeepers to student success. This project scales a proven “Targeted Mastery” pedagogical model in TMATH 109—where preliminary research established that engagement with remedial retakes is a statistically significant predictor of final exam performance (p=0.018). By implementing the AI-Driven Bridge System through interactive web-based platforms, we are developing automated tools to generate context-aware adaptive practice, track student performance trajectories and provide Socratic tutoring. This approach removes the logistical barriers to individualized remediation, transforming early mistakes into formative learning paths. The project provides a scalable, evidence-based framework designed to improve retention and pass rates in gateway quantitative reasoning courses.

Project team
Dinuka Sahabandu, Assistant Teaching Professor, Department of Electrical and Computer Engineering, College of Engineering
Luyao Niu, Assistant Teaching Professor, Department of Electrical and Computer Engineering, Global Innovation Exchange, College of Engineering
Payman Arabshahi, Professor, Department of Electrical and Computer Engineering, College of Engineering
Radha Poovendran, Professor, Department of Electrical and Computer Engineering, College of Engineering

Project summary
ENGINE×LAUNCH-AI is a multi-agent AI system designed to scale and enhance experiential learning in UW’s ENGINE capstone program and GIX Launch Studio. As these programs grow to serve over 300 students annually, manual processes for team formation, mentorship, progress tracking, and feedback become increasingly difficult to sustain. The proposed system introduces six AI agents to support key workflows including team formation, mentor recommendation, progress tracking, reporting, feedback synthesis and ABET-aligned evidence collection. The system operates with humans in the loop, ensuring faculty oversight and protecting sponsor confidentiality through structured, non-proprietary inputs. By automating coordination and improving visibility into project progress and risks, ENGINE×LAUNCH-AI aims to improve student outcomes, reduce instructional workload and enable scalable, consistent and high-quality capstone education across programs.

Project team
Cecilia Aragon, Professor, Department of Human Centered Design & Engineering, College of Engineering
Murtaza Ali, PhD Candidate, Department of Human Centered Design & Engineering, College of Engineering

Project summary
Data visualization literacy is increasingly vital in our data-centric world. The ability to understand, analyze, and create charts, graphs and other visual representations that compress information into human-comprehensible forms is essential for university graduates across all majors. However, creating effective visualizations typically requires both knowledge of visualization principles and programming skills to move beyond standard, template-based tools.

Current generative AI tools show promise in enabling non-programmers to harness programming flexibility for visualization tasks. Yet research examining how students use LLM assistance to create data visualizations remains limited. Through a mixed-methods approach combining quantitative and qualitative research, we will investigate how AI-assisted programming affects student learning outcomes in visualization creation. This study will illuminate not only effective pedagogical approaches for teaching data visualization with AI tools, but also the underlying mechanisms that make these pedagogies successful, informing curriculum design for all majors, not just technical ones.

Project team
Belinda Y. Louie, Professor, School of Education, UW Tacoma
Karlyn Davis-Welton, Instructor, School of Education, UW Tacoma
Janelle Franco, Instructor, School of Education, UW Tacoma
Jamie Lee, Instructor, School of Education, UW Tacoma
Elin Björling, Research Scientist, Department of Human-Centered Design & Engineering, UW Seattle

Project summary
Project TELL-AI is transforming how the University of Washington Tacoma prepares teachers for multilingual classrooms. While AI is rapidly changing education, many educators feel hesitant about its ethical use. Our project brings together four faculty members to systematically redesign ten courses in the Washington state English Language Learner endorsement licensing program. Using a structured iterative Design Thinking approach, we will move from identifying teacher needs to testing AI-enhanced lessons in actual classrooms. We aren’t building complex software; instead, we are creating a practical “blueprint” that shows teachers how to use AI as a collaborator for lesson planning and personalized student support. By modernizing this curriculum, we ensure that UWT graduates are ready to help K-12 students in high-need districts overcome language barriers. Ultimately, TELL-AI provides a scalable model for how any university department can move from individual experimentation to a unified, responsible program for the future of learning.

Project team
Katy E. Pearce, Associate Professor, Department of Communication
Mert Bayar, Postdoctoral Scholar, HCDE, Center for an Informed Public

Project summary
Over four years, the lead investigator has developed and refined a custom AI chatbot that provides tutoring feedback on conceptualization and operationalization – skills foundational to social science research methods but difficult to teach at scale. The tool has demonstrated measurable improvements in student work. However, it currently operates within ChatGPT, creating access barriers and storing student data outside UW systems. This project will migrate the tutor to UW’s Purple platform during Spring/Summer 2026, then evaluate it when the lead investigator teaches methods in Autumn 2026, and pilot it with another instructor to assess scalability. The project will also produce documentation enabling instructors across UW to adapt the tool for their courses. Research methods courses are offered in Communication, Sociology, Political Science, Psychology, Social Work, Public Health Information School and other units – potentially reaching thousands of students annually. This project will transform a successful single-classroom tool into a university-wide resource.

Project team
Xi Ma, Assistant Teaching Professor, Department of Asian Languages and Literature, College of Arts & Sciences

Project summary
This project develops an AI-assisted intensive reading curriculum for advanced Chinese learners. It integrates large language model tools and AI-supported media technologies to curate, adapt, and scaffold authentic contemporary Chinese texts and multimodal listening, viewing, and speaking materials, enabling students to engage critically with current Chinese discourse across text, audio and video in the AI era. Through structured reading, discussion, and performance-based tasks, students explore topics such as artificial intelligence, digital ethics, social change and media discourse while strengthening advanced Chinese literacy skills. The project will produce a reusable, Canvas-ready curriculum package for use across advanced Chinese courses.

Project team
JungHee Kim, Teaching Professor, Department of Asian Languages and Literature, College of Arts & Sciences
Yunee Kim, Lecturer, Department of Asian Languages and Literature, College of Arts & Sciences

Project summary
This project explores how generative AI can be used to support speaking practice in beginner level Korean language courses with large enrollments. In introductory language classes, students often have limited chances to speak and receive individualized feedback, which can reduce confidence and slow progress. This project uses AI tools to provide students with frequent, low-stakes speaking opportunities outside of class. Students complete short, guided speaking tasks and receive immediate AI feedback focused on overall communication rather than detailed error correction. Instructors then guide students in reflecting on this feedback and connecting it to course goals. By combining AI-supported practice with instructor oversight, the project aims to increase student confidence, participation and willingness to speak. The project also models a transparent and ethical approach to AI use in teaching and offers a scalable framework that can be adapted across languages and disciplines.

Project team
Brett W. Maurer, Associate Professor, Department of Civil & Environmental Engineering, College of Engineering

Project summary
Generative AI tools such as ChatGPT are now ubiquitous among engineering students, yet formal instruction on how to use these tools responsibly, critically, and effectively is entirely absent from engineering curricula. This project develops and evaluates a transferable, course-embedded AI literacy program that reframes generative AI as a fallible collaborator requiring human judgment rather than an answer-generating substitute. Implemented within a junior/senior engineering design course, the program uses AI-mediated revision cycles and reflective critique to deepen student understanding of engineering concepts, uncertainty and professional responsibility. The project requires no programming and relies on widely accessible large language models. Outcomes include a validated teaching protocol, assessment data on learning gains, and openly shareable instructional materials adaptable across engineering disciplines nationally.

Project team
Ekin Yasin, Teaching Professor, Department of Communication, College of Arts & Sciences
Matthew J Powers, Professor, Department of Communication, College of Arts & Sciences
Lara Bradshaw, Acting Instructor, Department of Communication, College of Arts & Sciences

Project summary
This study scales a proven AI literacy intervention across required graduate and undergraduate courses in the Department of Communication, serving 545+ students during autumn 2026 through spring 2027. Building on pilot data showing that 91.7% of graduate students became more strategic AI users after interventions, this project has students engage with AI tools through structured prompts, analyze exemplar versus non-exemplar transcripts, co-construct evaluative rubrics, and provide peer feedback, thereby developing critical thinking skills related to AI adoption. Pre- and post-assessments, focus groups and reflection exercises measure AI literacy gains, misinformation-detection capabilities and strategic application skills. A dedicated unit connects AI literacy frameworks to Communication’s media literacy tradition. The project also includes developing faculty workshops to optimize implementation and creating an open-source toolkit with prompt libraries, assessment instruments and facilitation guides for UW faculty. The study develops transferable resources that advance responsible AI integration while serving diverse populations.

Project team
Jennifer A Slyker, Associate Professor; Schools of Medicine and Public Health, Department of Global Health
Elizabeth Kirk, Teaching Professor; Department of Epidemiology: Food Systems, Nutrition and Health
Anya Nartker, Clinical Instructor and Managing Director of E-learning Program (eDGH); Department of Global Health
Alex McGee; Assistant IT Director; Department of Global Health
Christine Bastian, Assistant Director of Academic Programs & Student Services; Department of Global Health
Elizabeth Scott, MEd, Senior E-learning Developer, E-learning Program (eDGH); Department of Global Health
Chris Joss, Digital Media Specialist; Department of Global Health
Julie Nanavati: Senior E-Learning Developer, E-learning Program (eDGH); Department of Global Health

Project summary
GenAI is rapidly entering public health practice, creating urgent needs for workforce readiness and responsible use. We will develop a modular course on GenAI use for public and global health. Designed for graduate students and working professionals, the course will provide skills and frameworks for using GenAI effectively, safely and ethically in their work. The course will include applied projects. Cross-listed across the School of Public Health, the course will establish shared language and frameworks for GenAI use. The curriculum will focus on foundational competencies such as how GenAI works, prompt engineering, applied decision-making, ethics, equity, risk mitigation and accountability using UW-approved and open models. The course will prepare UW students while strengthening the domestic and global public health workforce. Leveraging existing e-learning infrastructure, this investment will produce an effective, up dateable and accessible course, adaptable for different learners with sustained impact across UW and beyond.

Project team
Michelle Koutnik, Associate Research Professor, Department of Earth and Space Sciences, College of the Environment

Project summary
Since change is fundamental not only to the Arctic climate and environment but also to societies that live in and interact with the Arctic, especially Indigenous Peoples and Arctic residents, Arctic Studies courses require integrated understanding on different topics and from different perspectives. This project will design, develop, and test an AI chatbot tool to support interdisciplinary learning about the Arctic. The tool will be trained on resources the students are reading to help them reason through interdisciplinary questions and challenge their assumptions. The tool will deliver credible sources on Arctic Indigenous perspectives and priorities while emphasizing data sovereignty and Indigenous Rights. A goal is that this chatbot tool will supplement student’s research and learning work, support how they generate ideas and advance in-class discussion that combine to inform their own critical thinking and project deliverables. The tool will be tested and reviewed with students and consulting scholars.

Project team
Emanuela Furfaro, Assistant Teaching Professor, Department of Statistics, College of Arts & Sciences

Project summary
LLteacher is an open-source web application that we have designed to help instructors integrate large language models (LLMs) into coursework in a structured, pedagogically aligned way. It guides students through interactive conversations that support, rather than replace, their reasoning, while giving instructors full visibility into the interaction process. This project will deploy LLteacher in STAT 311, a large introductory service course, to enhance learning, strengthen AI literacy and evaluate how structured LLM engagement affects student understanding. Funding will support the development of LLteacher specific homework, instructional materials for students and instructors, expanded data collection and a systematic evaluation of learning outcomes. The project will also produce reproducible curriculum materials and disseminate lessons learned to enable broader adoption of responsible, transparent AI supported homework design.

Project team
Sasha Seroy, Assistant Teaching Professor, School of Oceanography, College of the Environment
Katharine Bigham, Postdoctoral Research Scholar, School of Oceanography, College of the Environment
Ada Carter, Undergraduate Student, School of Oceanography, College of the Environment

Project summary
Computer vision AI can revolutionize ocean science research by enabling underwater imagery analysis at unprecedented scales, yet most marine scientists lack training to leverage these tools. We have successfully piloted a course teaching undergraduates to apply computer vision to authentic research problems, delivering it three times across two UW campuses since Winter 2025. However, a critical barrier remains: while the curriculum contains sufficient technical depth, it lacks comprehensive documentation needed for independent faculty adoption.

This project will transform our pilot into a broadly accessible educational resource by producing: (1) a fully documented instructor handbook with lesson plans and teaching notes; (2) a standardized quarter-long course package; (3) a condensed week-long intensive format; and (4) adaptation guides for other environmental disciplines. All materials will be openly licensed via oceancv.org, enabling sustainable scaling across UW and beyond while addressing workforce demand for environmental data scientists with integrated AI capabilities.

Project team
Chun Wang, Professor, College of Education
Min Li, Professor, College of Education
Yulia Tsvetkov, Associate Professor, Paul G. Allen School of Computer Science & Engineering, College of Engineering
David Arthur, Assistant Professor, Department of Sciences and Mathematics, UW Tacoma

Project summary
Each year, UW enrolls thousands of students in introductory courses delivered in large lecture formats. This standardized instructional approach, often implemented without adequate staffing or resources, is associated with failure rates of 30%–50%. Failure in these gateway courses can disrupt academic trajectories, leading students to switch majors or extend time to degree completion. Instructors of these courses often face the daunting task of generating, administering and scoring assessments to meet their teaching goals. This project will develop a “diagnostic assessment copilot” that supports instructors in rapidly creating high-quality assessment tasks aligned to course learning goals. The tool will be tested by faculty members from both Seattle and Tacoma in introductory statistics classes. By lowering barriers to high-quality assessment at scale, this project directly advances UW’s teaching mission and lays the groundwork for longer-term improvements in student success.

Project team
Menaka Abraham, Teaching Professor, School of Engineering & Technology, UW Tacoma
Andrea Hill, Teaching Professor, School of Social Work & Criminal Justice, UW Tacoma
Lisa Hoffman, Professor, School of Urban Studies, UW Tacoma
Julia Eaton, Associate Professor, School of Interdisciplinary Arts & Sciences, UW Tacoma
Sean Schmidt, Executive Administrator, Student Planning and Administration, UW Tacoma

Project summary
Our project performs the urgent task of building AI literacy among UWT’s largely first-generation, transfer, and working student population as they enter an AI-saturated job market and world. Our innovative approach consists of two elements: I) A “bookended” AI literacy learning experience consisting of a broader “this is the AI world” 2-credit seminar for new students, and deeper 2-credit senior capstone in which seniors bring their accumulated disciplinary expertise to examine AI in their professional contexts, futures, and beyond, and II) A host of modifiable “plug and play” assignments and activities for faculty from any discipline to use without significant investment in AI development. Importantly, this approach will enhance AI literacy over the course of the UWT experience by focusing on our student population’s lived experiences. We envision that our campus becomes a model for AI literacy education serving non-technical student populations, aligning with our mission to serve the region.

Project team
Lorne Arnold, Assistant Professor, Civil Engineering, School of Engineering & Technology, UW Tacoma
Chris Marriott, Teaching Professor, Computer Science and Systems, School of Engineering & Technology, UW Tacoma
Cassandra Donatelli, Assistant Professor, Mechanical Engineering, School of Engineering & Technology, UW Tacoma

Project summary
AI tools offer an opportunity to reduce non-essential struggles for students learning to code (like fixing syntax errors), freeing students to focus on deeper learning. But they also pose a challenge to learning because typical AI agents are designed to increase user productivity, not necessarily user understanding. Drawing on Polya’s four-step problem-solving framework (understand, plan, execute, look back) and recent research on cognitive engagement techniques, we will develop an AI teaching assistant that guides students through structured inquiry and problem decomposition rather than providing direct solutions. We will deploy this system across civil/mechanical engineering and computer science courses, comparing learning outcomes to baseline conditions. Our goal is to help students develop genuine problem-solving skills while still benefiting from AI assistance with syntax and implementation details.

Project team
R. Sean Bethune, Instructor, UW Professional and Continuing Education

Project summary
This project updates and modernizes the software engineering curriculum at UW Continuum College shifting from “copilot” AI tools to autonomous “agentic” systems. Currently, students use AI as a simple assistant for writing code. However, the technology industry increasingly requires engineers who can manage AI “agents”—autonomous digital coworkers that can independently plan, test and fix software modules. We will develop a new curricular framework that teaches professional learners how to design, supervise and set safety guardrails for these autonomous systems. By training students to act as orchestrators of AI agents rather than just users of AI tools, we ensure they are prepared for the high-level leadership and technical roles demanded by Seattle’s leading technology firms. This project will serve over 300 professional learners annually and create a “Faculty Toolkit” of open-access resources to help other UW departments integrate agentic AI into their specific disciplines.

Project team
Lilo D. Pozzo, Professor, Department of Chemical Engineering and Department of Human Centered Design & Engineering, College of Engineering
Nada Y. Naser, Assistant Teaching Professor, Department of Chemical Engineering, College of Engineering

Project summary
The Department of Chemical Engineering offers two required materials laboratory courses focusing on Colloid and Interface Science (ChemE 455) and on Polymer Materials Synthesis and Thermodynamics (ChemE 460) that all students must take. The proposed program will transform legacy laboratory modules to introduce high-throughput experimentation via laboratory automation (e.g., pipetting robotics and analytics) and integration of agentic artificial intelligence (AI) methods to control and interpret large datasets and associated experimental outcomes. We will use funds to support the creation of new modules that teach the same foundational materials science principles (i.e., polymers, colloids, and surfaces) while also introducing principles of AI, including experimental design, Bayesian optimization and active learning. The transformed courses will better prepare students for modern careers in industry and research where AI-driven automation is rapidly becoming standard practice.

Project team
Bo Zhao, Professor, Department of Geography, College of Arts & Sciences

Project summary
This project will transition an existing special topics Geospatial AI course into a full-quarter, institution-ready course focused on AI, social good, and environmental sustainability. The course combines hands-on geospatial AI workflows (machine learning, deep learning, and large language models) with critical social theories to examine housing inequality, homelessness, algorithmic bias and the environmental footprint of AI. All labs run in Google Colab and are supported by detailed technical documentation and critical readings. Course materials are already openly published on GitHub and have attracted international users, including translations, indicating broad reuse potential. SEED-AI funding will support curriculum expansion, lab optimization, integration of fast-moving AI scholarship and emerging practices (e.g. vibe coding, etc.), development of rubrics and learning assessments and creation of instructor-ready materials for reuse across UW and beyond.

Project team
Kelsey M. Conrick, Postdoctoral Scholar, Center for Firearm Injury Prevention, Department of Pediatrics, School of Medicine
Anna Bender, Postdoctoral Scholar, Department of Pediatrics, School of Medicine
Nova Rivera, Health Services Program Manager, Department of Rehabilitation, School of Medicine
Megan Moore, Sidney Miller Endowed Associate Professor of Direct Practice, School of Social Work
Ali Rowhani-Rahbar, Bartley Dobb Endowed Professor, Department of Epidemiology, School of Public Health
Stacey De Fries, Associate Teaching Professor, School of Social Work
Sarah F. Porter, PhD Candidate, School of Social Work
Adam Davis, PhD Candidate, School of Social Work
Erika Marts, Program Manager, Department of Epidemiology, School of Public Health

Project summary
This project will bring Pathways 2 Safety (P2S) to UW social work (undergraduate and graduate) nursing (graduate) students. P2S is a 3-hour interdisciplinary training to teach clinicians to have firearm safety conversations with clients. After the training, students will be randomized to practice counseling skills using either AI-based clients or trained standardized clients (i.e., live actor) to compare the two practice modalities.

AI-based clients include a coaching agent that provides individualized feedback. Standardized client encounters will have parallel debriefing protocols. We will have three main evaluation outcomes: (1) effectiveness through pre/post surveys and observed standardized client assessments, (2) modality acceptability based on student engagement and feedback and (3) pragmatic comparison of modality time and cost.

Because participation is voluntary, this project provides a low-stakes opportunity to examine the educational value and acceptability of AI-supported learning relative to traditional methods, generating evidence to inform responsible, scalable use of AI across UW.

Project team
Gaj Sivandran, Assistant Teaching Professor, School of Environmental and Forest Sciences, Center for Quantitative Sciences, College of the Environment
Kat Huybers, Assistant Teaching Professor, Department of Atmospheric and Climate Sciences, College of the Environment
Kerry Naish, Professor, Director of Marine Biology, College of the Environment
Mikelle Nuwer, Teaching Professor, Associate Director for Undergraduate Programs in School of Oceanography, College of the Environment
Kristi Straus, Teaching Professor, Associate Director of Program on the Environment, College of the Environment
Jane Dolliver, Professional Staff, College of the Environment
Akshay Mehra, Assistant Professor, Department of Earth and Space Sciences, College of the Environment

Project summary
Generative AI has the potential to enhance undergraduate learning, but only when its use is treated as a pedagogical design challenge rather than a policy or enforcement issue. This project advances the University of Washington’s educational mission by piloting learning-centered approaches to AI integration within the College of the Environment. By pairing a student-facing course with faculty-led pedagogical development, the project explores how AI can support learning, discovery and ethical reasoning without displacing core intellectual work. Faculty Learning Communities will develop and share concise teaching cases that model responsible, learning-aligned AI use across environmental contexts. In parallel, a new 200-level undergraduate course, first offered in Spring 2026, will teach students to engage with AI as a learning tool through active, reflective practice. Together, these efforts will generate transferable resources, clarify expectations for appropriate AI use and provide scalable models for responsible AI-informed pedagogy across disciplines at UW.

Project team
Jennifer Mankoff, Professor, Paul G. Allen School of Computer Science & Engineering, College of Engineering
Brianna Wimer, Visiting Researcher, Paul G. Allen School of Computer Science & Engineering, College of Engineering

Project summary
Instructional diagrams are widely used across university courses to explain core concepts, processes and algorithms, yet they are typically shared as static images that are inaccessible to blind and low-vision students. Creating accessible versions of these diagrams currently requires significant expertise and instructor time. This project develops and pilots AccessFlow, an AI-powered tool that converts node-link diagram images into accessible representations students can explore independently, and that instructors can easily integrate into courses or share with others. We will evaluate the prototype with Summer 2026 instructors on time savings, ALT text completeness, and accuracy of AI-generated results, then deploy it across UW in Fall 2026. This will help UW meet federal accessibility requirements, allow us to collect a dataset of diagrams and remediations to improve the system, support inclusive teaching by reducing barriers for disabled students and create reusable instructional resources for faculty across disciplines.

Project team
James Weichert, Assistant Teaching Professor, Paul G. Allen School of Computer Science & Engineering, College of Engineering
Megan Hazen, Associate Teaching Professor, Paul G. Allen School of Computer Science & Engineering, College of Engineering

Project summary
This project, led by Allen School of Computer Science & Engineering professors James Weichert and Megan Hazen, involves the development of an introductory ‘AI literacy’ course on the “Principles, Applications and Impacts of Artificial Intelligence” open to all undergraduate students at UW. The course will provide a broad introduction to AI, preparing students from a wide range of backgrounds and disciplines to engage critically in an increasingly AI-integrated world. The primary focus of this course is the critical evaluation of AI systems by (1) developing an understanding of the core computing principles that enable AI; (2) exploring the uses and limitations of current AI models through guided experimentation; and (3) discussing the impacts of AI across diverse domains.

SEED-AI grant funding will support the cost of the course’s development in summer 2026 for a first pilot offering in winter 2027 under a CSE 190 special topics designation.

Project team
Kaylea Champion, Assistant Professor, Computing and Software Systems, School of STEM, UW Bothell
Jeffrey Kim, Associate Teaching Professor, Computing and Software Systems, School of STEM, UW Bothell

Project summary
This project focuses on teaching students the impacts to a software engineering workflow – including planning and estimation, requirements analysis, design, development, testing, deployment and maintenance – when AI-powered features are part of the product. Although substantial attention has been paid to AI as a code generating tool and to the technical skills involved in building AI models, impacts of AI features on overall engineering workflow is relatively less explored. All computing students should have an awareness of these process impacts, regardless of their specialties and goals. We propose to develop and evaluate two interventions: a no-code hands-on simulation of evaluating and fine-tuning a model, and a series of enhancements and extensions for learning modules focused on different stages of a development project. We propose to develop a preliminary version of these interventions, then to evaluate and refine them through a series of co-design workshops, and finally to publish our results.

Project team
Sarah J. Iribarren, Associate Professor, Biobehavioral Nursing and Health Informatics, School of Nursing
Andrea Hartzler, Professor, Biomedical Informatics and Medical Education, School of Medicine
Michael Leu, Professor, Pediatrics and Biomedical Informatics and Medical Education, School of Medicine
Weichao Yuwen, Associate Professor, School of Nursing & Healthcare Leadership, UW Tacoma
Jan Flowers, Sr. Research Scientist, Biobehavioral Nursing and Health Informatics, School of Nursing
Emily Schildt, Clinical Informatics Fellow, Biomedical Informatics and Medical Education, School of Medicine
Sikha Pentyala, Postdoctoral Researcher, UW Tacoma School of Engineering and Technology
Kelly Schorling Brewer, Assistant Teaching Professor, Biobehavioral Nursing and Health Informatics, School of Nursing
Jennifer Sprecher, Project Manager, Biobehavioral Nursing and Health Informatics, School of Nursing
Patricia Reid Ponte, Associate Affiliate Clinical Professor, Biobehavioral Nursing and Health Informatics, School of Nursing
Jared Erwin, Lecturer, School of Medicine, Department of Biomedical Informatics and Medical Education

Project summary
Artificial intelligence (AI) is increasingly embedded in healthcare systems, but most healthcare professionals have received little formal training in its responsible clinical use. This project will develop short, modular learning resources (10–15 minutes each) to prepare healthcare professionals in the UW Clinical Informatics and Patient-Centered Technology (CIPCT) graduate program to understand, evaluate, and ethically apply AI-based tools in clinical settings. The modules will be developed with input from learners, clinical informatics trainees, faculty and clinical leaders to ensure clinical relevance and alignment with real-world practice. An AI-supported tutor will complement the modules by allowing learners to engage in interactive, scenario-based practice focused on ethical decision-making and clinical judgment. Project success will be evaluated through brief knowledge assessments, learner feedback and usability evaluations. While designed for CIPCT, these resources will be readily adaptable for other healthcare programs and clinical partner settings.

Project team
Kevin Lin, Assistant Teaching Professor, Paul G. Allen School of Computer Science & Engineering, College of Engineering
Chris Holstrom, Assistant Teaching Professor, Department of English, College of Arts & Sciences
Ben Lee, Assistant Professor, Information School
Kathryn Pursch Cornforth, Director of Community Engagement, Community Engagement and Leadership Education (CELE) Center
Suh Young Choi, Part-Time Lecturer, Paul G. Allen School of Computer Science & Engineering, College of Engineering
Megumi Kivuva, PhD Candidate, Information School

Project summary
We propose developing, deploying, evaluating and sharing a framework for AI-accelerated community-engaged learning (CEL) in multiple undergraduate computing courses at UW using an instructor-led approach during Autumn 2026. The instructor would collaborate with community partners to curate broad research questions present in their work and then use disciplinary expertise to focus those questions toward specific instructions for students. This approach aligns course learning outcomes with CEL activities, presents opportunities for realistic integration of AI tools and addresses student desire for projects with real-world application. Early conversations with community contacts have identified viable research questions, such as the application of existing open data that could support city planning or small businesses in specific communities. This initial investment to develop these CEL activities can then be reused in future offerings and for course contexts beyond the courses in the Autumn 2026 pilot.

Project team
Karen Cheng, Professor, Visual Communication Design, College of Arts & Sciences
Li-Yuan Chiou, M.Design Candidate, Interaction Design, College of Arts & Sciences

Project summary
Design education centers on critique sessions where students present work and engage in dialogue with peers and faculty to develop visual literacy. However, critique is fundamentally resistant to scale. In courses with 60-100+ students, critique occurs only once weekly, leaving students without feedback during iterative work. We propose developing an AI-powered critique tool that evaluates formal composition through visual annotation rather than text-only feedback. Students upload images and receive annotated visuals: arrows indicating visual flow, markers highlighting focal points, overlays revealing symmetry axes and diagrams identifying contrast areas. The tool assesses asymmetry, unity/variety balance, visual flow and spatial depth, providing 2-3 distinct revision suggestions per issue that require students to exercise design judgment. We validate through iterative testing with 20+ student examples (targeting 80% accuracy), then deploy as an opt-in resource in a sophomore design course. Students can share annotated visuals to discuss with peers and faculty, enabling both individual AI critique and collaborative learning.

Project team
David Arthur, Assistant Professor of Statistics, School of Interdisciplinary Arts & Sciences, Department of Sciences and Mathematics, UW Tacoma
Zaher Kmail, Associate Professor of Statistics, School of Interdisciplinary Arts & Sciences, Department of Sciences and Mathematics, UW Tacoma

Project summary
Project- and problem-based learning activities have been shown to improve learning outcomes for students. In introductory statistics courses, such activities require students to analyze real-life datasets. To achieve the best outcomes, each student would be provided with a personalized dataset that aligns with their interests and goals while still facilitating progress towards specific learning objectives. In practice, finding real datasets can be difficult and creating high-quality, realistic synthetic datasets can be time-consuming. This project proposes the use of Large Language Models (LLMs) to realize the ideal of personalized datasets for classroom learning. We propose using LLMs to create AI agents that perform the tasks required to write code that can be used to produce realistic, synthetic datasets. This tool will be tested and validated by both instructors of, and students in, an introductory statistics course with the ultimate goal of improving engagement and learning outcomes for students.

Project team
Louisa Mackenzie, Associate Professor, Comparative History of Ideas, College of Arts & Sciences

Project summary
This project advances a humanistic response to AI adoption in higher education through two deliverables: a scalable general education course in Critical AI Literacy (CAIL), and resources for instructors. CAIL starts from the principle that knowing when and why *not* to use AI tools is as important as knowing how to use them. It presents non-use both as informed method and object of inquiry. Students examine how AI reshapes individual and collective knowledge, AI’s social impacts and what is lost or gained when thinking is offloaded to synthetic generation, thus making more informed choices about their own (non) use. CAIL is in dialogue with, not opposed to, AI adoption. It is contingent, not absolute, while respecting principled refusal as a legitimate intellectual stance. The project also develops resources for instructors who, for evidence-based pedagogical reasons, seek to prioritize minimally mediated, human centered teaching and assessment in an AI forward environment.

Project team
James Fogarty, Professor, Paul G. Allen School of Computer Science & Engineering, College of Engineering
Jesse Martinez, PhD Student, Paul G. Allen School of Computer Science & Engineering, College of Engineering

Project summary
We propose to design, build, and pilot an exploratory tool that provides AI-based feedback aimed at helping students avoid common pitfalls in courses that emphasize learning through open-ended design-based project sequences. We will initially focus on pitfalls in the project sequence of CSE 440: Introduction to Human-Computer Interaction, noting its project sequence shares important properties with many other UW courses. Across such courses, failures in early design artifacts often propagate into later assignments, making feedback less effective and more costly for both students and course staff. We propose an approach where assignment-specific AI-based feedback is configured by instructors, emphasizing existing learning goals and anchored in pitfalls instructors already articulate. AI-based feedback thus aims to complement feedback from course staff, providing always-available feedback on common pitfalls while allowing course staff to focus limited resources on more nuanced feedback around student creativity.

Project team
Narjes Abbasabadi, Assistant Professor, Department of Architecture, College of Built Environments

Project summary
Artificial intelligence (AI) offers powerful new ways to support design reasoning, exploration, and data-informed decision-making, yet these capabilities remain largely untapped in architectural education. This project addresses a critical dual-literacy gap, as current curricula lack both technical fluency across AI methods and critical literacy in ethics, agency and societal impact. DesignAI develops and implements an AI-augmented educational and research framework that repositions AI beyond automation as a scaffold for design reasoning within design workflows, advancing problem framing, iterative exploration and decision-making. The framework integrates a shared foundation in AI-driven design reasoning with project-based pathways, enabling students to develop depth in selected methods (e.g., machine learning, including deep, physics-informed, and multimodal approaches, as well as generative and agentic AI systems) aligned with their research questions. Implemented in seminars and research studios focused on computational and performance-driven design, the project evaluates impacts using measurable indicators of design reasoning quality, decision-making under uncertainty and ethical and critical awareness, generating generalizable insights into AI-mediated design cognition and human–AI interaction, and producing adaptable curricular models for broader adoption.

Project team
Christopher R. Beasley, Associate Professor, Department of Social Sciences, School of Interdisciplinary Arts & Sciences, UW Tacoma

Project summary
Faculty resources for teaching with generative AI typically present isolated use cases without showing how applications connect into a coherent workflow. This project addresses that gap by developing a comprehensive faculty AI workflow guide drawn from three years of iterative development across synchronous and asynchronous courses. The guide will document how generative AI integrates with broader digital tools to create a teaching production pipeline grounded in backward design and constructive alignment. It will include visual workflow diagrams, tested prompt templates, decision frameworks for when AI adds value and strategies for maintaining instructor voice and pedagogical intentionality. Rather than disconnected tips, faculty will gain a framework where AI-assisted outputs build on one another from course design through feedback — capturing the efficiency and coherence that emerge when AI integration becomes a workflow rather than an add-on. A small pilot group of faculty from diverse disciplines will validate transferability before broader dissemination to UW faculty and the public.

Project team
Léonard Boussioux, Assistant Professor, Michael G. Foster School of Business, Department of Information Systems and Operations Management
Rebekah Lee Baik, Doctoral Student, Michael G. Foster School of Business, Department of Information Systems and Operations Management
Caitlin Cunningham, Doctoral Student, Michael G. Foster School of Business, Department of Information Systems and Operations Management
Archit Gupta, Master of Science in Business Analytics Candidate, Michael G. Foster School of Business
Anushka Mathur, Master of Science in Business Analytics Candidate, Michael G. Foster School of Business
Leela Nageswaran, Assistant Professor, Michael G. Foster School of Business, Department of Information Systems and Operations Management

Project summary
This project develops three interconnected, open-source resources that teach students to build, evaluate and improve AI systems through hands-on practice. First, vibe coding tutorials guide students with no programming background to create portfolio-worthy websites, apps, and autonomous agents using conversational AI interfaces. Second, Agentic Cases—a novel pedagogical format—provide pre-coded AI systems that students probe, test and improve through direct experimentation, mirroring professional AI evaluation workflows. Third, an open-source video analysis research tool automatically analyzes screen recordings of student-AI interactions, enabling rigorous empirical study of how learners engage with AI tools. All materials are production-ready, openly shared, and aligned with the Foster School’s AI Learning Outcomes. Built on three years of field-tested pedagogy across six AI courses and fifteen boot camps reaching 1,500+ students and industry practitioners, the project delivers scalable infrastructure for teaching and studying AI-enhanced learning across UW and beyond.

Project team
Carl T. Bergstrom, Professor, Department of Biology, College of Arts & Sciences
Jevin West, Professor, Information School

Project summary
Large language models (LLMs) have upended education. Students and faculty alike are struggling to understand what it means to be a learner, scholar and human being in a ChatGPT world. We propose to develop, teach and disseminate, worldwide, a general education course, Modern Day Oracles or Bullshit Machines, for every college freshman and high school student wanting to reflect on what it means to be human in an LLM-infused world. The course takes a collaborative learning approach to a fundamental question grounded in the humanities: how can we learn and thrive with LLMs? In addition to teaching the course at the UW, we detail an ambitious yet demonstrably achievable plan to disseminate our curriculum worldwide as we have successfully done with previous curricula: through free and engaging online learning materials and extensive outreach to instructors at all levels, around the globe and across the disciplinary landscape.