AI & Society Research Center

Adaptive AI: How Can AI Alleviate Healthcare Workforce Shortages?

Can artificial intelligence help address the growing shortage of healthcare workers? 

Professor Cheon's project investigates how hospitals are using AI to support staffing, scheduling and administrative decision-making, and whether these technologies improve workforce capacity while reducing inequities across healthcare systems. 

The findings will help inform policies that promote more effective and equitable AI adoption in hospitals nationwide. 

Does AI Erode the Nuclear Revolution?

Artificial intelligence is transforming military technologies, but could it also undermine global nuclear stability? 

Professor Clary examines how AI may affect nuclear deterrence by improving surveillance, target identification and military decision-making. 

His research explores whether these advances increase the risk of conflict and what they mean for international security in an AI-driven world. 

Is This Image Yours?

What happens when AI mistakes its own creations for historical photographs? 

Professor Holmes combines AI image generation with traditional analog photography to explore how machines interpret visual history, memory and authorship. 

By documenting AI's surprisingly confident, but often inaccurate, explanations of images it originally created, the project raises important questions about truth, creativity and the future of visual culture. 

Developing an AI-Enhanced, Identity-Based Intervention to Promote Prosocial Development Among Justice-Involved Youth

Many young people involved in the justice system receive little support despite remaining at risk of future involvement.

Professor Jian is designing an AI-assisted intervention that helps youth imagine positive futures, strengthen their sense of identity and build motivation for lasting change. 

The project explores how AI can complement — not replace — human mentorship in supporting rehabilitation and reducing recidivism. 

Errant Hands

This project transforms one of AI image generation's most recognizable mistakes into a creative exploration of human embodiment, perception and machine vision. 

Through sculpture, 3D printing and AI-generated imagery, Professor Lawrence examines what these "errors" reveal about the limits of artificial intelligence and the complexity of the human body. 

The resulting exhibition invites audiences to reflect on how AI sees, and sometimes misinterprets, the world around us. 

Trustworthy AI for Multi-Scale Wildfire Smoke Prediction and Indoor Health Protection

Wildfire smoke is becoming an increasingly serious public health threat, even far from the fires themselves. 

Professor Li is developing AI tools that predict how outdoor smoke affects indoor air quality in schools, workplaces and public buildings. 

By combining environmental data with building information, the project aims to help communities make safer decisions during smoke events. 

Causal Machine Learning and Alcohol Use Trajectories: Understanding Cardiovascular Risk

Cardiovascular disease affects communities differently, but we still know little about how lifelong patterns of alcohol use shape these disparities. 

Professor Masum uses advanced causal machine learning methods to examine how alcohol consumption across the life course influences heart health among different racial, ethnic and socioeconomic groups. 

The project aims to generate evidence that supports more equitable approaches to disease prevention and public health. 

Do LLMs Constrain or Expand the Scope of Scientific Search & Discovery?

Large language models are changing how researchers search for scientific knowledge, but do they encourage discovery or narrow our perspectives? 

Professor Murray's project examines whether AI-powered search tools expose researchers to more diverse ideas or reinforce familiar patterns. 

The research will help guide the design of AI systems that better support creativity, innovation and scientific discovery. 

Addressing School Bullying with AI-Enhanced Professional Development for K-12 Educators

Teachers often have only seconds to respond when bullying occurs. 

This project develops an AI-powered training platform that allows educators to practice difficult conversations with virtual students in a realistic, low-risk environment. 

By providing immediate feedback, the platform aims to strengthen teachers' communication skills, confidence and empathy, ultimately helping schools create safer and more supportive learning environments. 

Digital Surveillance and Targeted Repression in Hybrid Regimes: Evidence from Pakistan

Governments around the world are increasingly using AI-enabled surveillance technologies, but what are the consequences for democracy and civil liberties? 

Professor Siddiqui investigates how digital monitoring tools are used to target political opponents and minority communities in Pakistan. 

The project will generate new insights into AI, governance and human rights in countries where democratic and authoritarian institutions coexist.

Upskilling to Assess LLM-Based Automated Content Analysis

Large language models are changing how researchers analyze massive collections of text, but are they more reliable than existing methods? 

Professor Young compares AI-powered content analysis with traditional rule-based approaches to determine when these new tools improve research and when transparency and reproducibility remain essential. 

The findings will help researchers make informed choices about AI-assisted analysis. 

Investigating Older Adults' Prompting Behavior with LLMs in Disaster Preparedness

As more people turn to AI for information, how can these tools better support older adults during emergencies? 

Professor Yuan's research explores how older adults interact with AI systems such as ChatGPT when preparing for disasters.

By understanding the challenges they face, the project seeks to improve AI tools so they are more accessible, trustworthy and useful during times of crisis.