Clark University has announced the recipients of the inaugural AI Innovation Grants, designed to support creative and impactful projects that explore the use of artificial intelligence in teaching, research, scholarship, institutional operations, or enhancements to learning, convening, and campus life.
The grants are supported by the Clark AI Innovation Fund, established with a $50,000 gift from the Yee Family, including Trustee Brian Yee ’93.
“These grants aim to foster a deeper understanding of AI and its potential as a tool to advance our academic mission, improve institutional effectiveness, and enhance the way we work with each other or with members of the Worcester community, among other objectives,” President David Fithian and Provost John Magee wrote in a campuswide email announcing the fund.
The selection committee — Vice President for Information Technology Joseph Kalinowski, President David Fithian, Dean of Research Jennifer Hanselman, and Professor Kat Andler — reviewed 59 grant applications, of which eight were chosen. Awardees include undergraduate and graduate students, faculty, and staff from across campus, including the departments of Computer Science, Political Science, and Psychology, the Graduate School of Geography, and the Becker School of Design & Technology.
The awardees must complete their work by August 2026 and share the results with the Clark community.
AI Innovation Grant Recipients
Lauren Gallagher ’26, MFA ’27, interactive media major (BSDT)
Intelligent NPCs: Exploring Generative AI for Adaptive Character Design in Game Development
Non-player characters (NPCs) are central to how players experience story, emotion, and agency in video games, yet most remain static, repeating predetermined lines regardless of context. As generative AI reshapes creative practices, very few games have bridged creative writing, machine learning, and player interaction to create believable, adaptive characters.
This project builds on Gallagher’s current MFA research in interactive media and game design, where she has been developing an Unreal Engine prototype that connects to Ollama, a local language model, and has discovered the transformative potential of generative AI in interactive storytelling to change the way players connect with virtual characters.
Antonio Galvao da Fonseca, Ph.D. student in geography; Professor Robert “Gil” Pontius, geography
Scaling Land Cover Change Analysis: Agentic AI for Global Environmental Monitoring
A crucial component of global environmental change is land change, which remains a major source of funding for the Clark Center for Geospatial Analytics. Petabytes (equal to one quadrillion) of classified land cover data exist globally, yet scientists analyze only regional extents due to computational limitations.
In the journal Transactions in GIS, Pontius and da Fonseca published a novel method that decomposes temporal changes into five components to rapidly identify the classes and time intervals that account for important changes. This project applies Agentic AI to automate large-scale computation of land cover change metrics using land use and land cover time-series maps. The system autonomously processes millions of pixels, calculates five change components, and generates analytical visualizations without continuous human intervention.
Professor Hyungsin Kim, computer science
SocraTalk: Learning by Doing, Reflecting by Design — Using Socratic AI to Unmask Dark Patterns
Artificial intelligence has quietly become one of the most powerful forces shaping how students learn, communicate, and make choices. Yet many first- and second-year undergraduates remain unaware of how AI-driven interface design and manipulative user experiences, often called dark patterns, shape their attention, emotions, and autonomy.
This project helps prepare students to use AI tools, question them, and, most critically, interpret them, and introduces an experiential learning model to cultivate critical awareness of digital manipulation. Kim and her students will design, implement, and evaluate “SocraTalk,” a Socratic-style AI learning environment embedded within a simulated app that reproduces common dark patterns, such as infinite scrolling, pre-selected defaults, and social-proof cues. Through guided AI dialogue, students will experience manipulation firsthand and articulate their reasoning and emotions about it.
Professor Shuo Niu, computer science; Jimin Lee, Ph.D. student, psychology; Torin Anderson ’27, computer science major
Developing and Teaching Meta-Prompting for Media Production
This project aims to develop a tool to support meta-prompting — an advanced generative AI (GenAI) prompting technique that emphasizes the structural and syntactical composition of media content. Building upon an ongoing project that reconfigures lecture videos into alternative explanations, Professor Niu seeks to design a meta-prompting framework that enables instructors and students to use GenAI to create new short-form educational videos.
The main objectives of this research are to identify the meta-information necessary to construct effective GenAI prompts in educational video creation and to teach students how to adopt the meta-prompting technique as an advanced skill for producing media content using GenAI.
Abraham Rahman ’27, psychology major; Preeti Bachu ’26, psychology major; Professor Michael Miller, psychology
AI-Generated Summaries of Complex Scientific Articles: Supporting Emotion Regulation in ADHD Readers
This study builds on previous findings from an undergraduate research study at the University of Michigan Flint, which evaluated the potential of large language models to create accessible summaries of scientific articles for individuals with ADHD and Autism Spectrum Disorder.
This project aims to generate quantitative evidence to inform future AI summary design choices that better accommodate emotional regulation-based challenges (e.g., frustration, difficulty shifting attention) for people with ADHD, making summaries more accessible by design. Also collaborating on this project are Skyler Koba, a senior at Pennsylvania State University, and Steve Wilson, a computer science professor at the University of Michigan–Flint.
Onyx Rothman ’26, political science major
Bridging the Digital Divide: Developing Inclusive AI Transcription Technology for Worcester’s Diverse Communities
Current AI transcription services are influenced by historical inequalities embedded in their training data, based on internet content that skews overwhelmingly white and Western — tech companies prioritize scale over specificity, training models on massive datasets that flatten linguistic diversity into standardized forms, leading to results that systematically misrepresent speakers of African American Vernacular English and other non-standard dialects.
This project builds upon TranscribeWell, an existing operational transcription platform that already demonstrates superior contextual understanding compared to mainstream services. Rothman is collaborating with data scientist Aaron Bergman to develop a specialized transcription model trained specifically on Worcester’s linguistic diversity, particularly the city’s large population of people from Ghana, in hopes of demonstrating that localized AI development is feasible and can preserve rather than erase cultural and linguistic identity.
Sarina Talerico ’27, computer science major; Solan Homestead ’27, computer science major
AI Trashcan Sorter: AI in Improving Sustainability and Waste Management at Clark University
Effective waste management is important for environmental sustainability, yet misclassification of recyclable, compostable, and landfill waste is a challenge on Clark’s campus, as students don’t have time to properly dispose of their waste or simply ignore bin labels. This results in significant volumes of mixed waste, which in turn contaminates other sustainable efforts. This project involves developing an AI-powered trash-sorting system that classifies and redirects waste to the appropriate disposal streams. Using sensors, cameras, and a sorting algorithm, Talerico and Homestead’s plan will have both immediate use and create a scalable prototype for widespread implementation across campus. The system will combine robotics, AI, and real-time feedback to reduce misclassification while also collecting data on waste patterns.
Evan Wilson, associate dean of the college; Laurie Ross, dean of the college; Mark Jacobs, professor in the master’s in AI program; Center for Excellence in Teaching, Assessment, and Learning Steering Committee
Supporting Faculty to Better Understand and Utilize Artificial Intelligence for Teaching and Learning
This project will help faculty understand and integrate AI into their teaching, enabling them to better prepare students for a future in which knowledge and AI skills will be essential to success in every field. Embracing AI thoughtfully means using it to enhance, and not replace, relationships, creativity, and critical thinking. A collaborative team led by faculty and staff from the Dean of the College and the Center for Excellence in Teaching, Assessment, and Learning seeks to build faculty capacity and confidence with AI tools to support teaching and learning, ensuring that Clark’s classrooms are sites of innovation, ethical inquiry, and human connection in an AI-shaped world. This work will begin with a series of faculty workshops during the spring and summer of 2026.
