AI research earns triple honors

Assistant Professor of Information Systems Tian Lu recognized with three INFORMS awards for groundbreaking research in AI, data analytics, and social responsibility.

Every fall, the Institute for Operations Research and the Management Sciences (INFORMS) recognizes academics, organizations, and institutions of higher education that demonstrate excellence in operations research and analytics. This year, Assistant Professor of Information Systems Tian Lu is the recipient of three significant INFORMS awards:

• The INFORMS Information Systems Society (ISS) Gordon B. Davis Young Scholar Award.
• The INFORMS eBusiness Section Best Paper Award for his paper "From Signal to Noise: How Widespread LLM Usage Transforms Evaluator Behavior in Credit Screening.
• The ISS Bapna-Ghose Social Justice Best Paper Award for his paper "Profit vs. Equality? The Case of Financial Risk Assessment and A New Perspective on Alternative Data."

Lu, a member of INFORMS since his PhD studies, focuses on impactful research that explores data-driven decision-making, emerging information technology and AI trends, behavioral theories, and the value of data in balancing economic benefits with social welfare.

"I am deeply grateful to the award committees for their acknowledgement, which serves as a testimony to the value of our collective efforts and strengthens my motivation to keep learning and contributing," says Lu. "These awards recognize my contributions in the direction of exploring how to navigate AI mindfully and reaffirm my mission to generate knowledge that advances both economic efficiency and social welfare."

Lu's paper, "From Signal to Noise: How Widespread LLM Usage Transforms Evaluator Behavior in Credit Screening," examines how extensive use of large language models (LLMs) by applicants reshapes evaluator behavior in credit screening. The study reveals that when applicants rely widely on LLMs, their applications become less informative and stylistically similar, influencing how evaluators make decisions, and such negative impacts could also spill over to applicants who did not use LLMs in their submissions."

"The study highlights a broader paradox of AI democratization: tools that enhance communication at the individual level can, when used too widely, degrade collective decision quality in information-intensive evaluations," Lu explains.

His second paper, "Profit vs. Equality? The Case of Financial Risk Assessment and A New Perspective on Alternative Data", published in MIS Quarterly in 2023, investigates how alternative data — such as information from smartphones, social media, and online shopping — can improve both financial inclusion and business profitability in credit risk assessment. The research found that behavioral data from smartphones is the most effective, improving predictive performance and expanding access to credit for underserved populations, while online shopping data tended to reinforce inequality due to the consumers’ socioeconomic status.

"These insights advance the broader goal of algorithmic fairness and social justice, showing how responsible data design can align commercial success with equitable access in financial decision-making," says Lu.

As reviewer for several leading INFORMS journals, including Management Science, Information Systems Research, MIS Quarterly, Manufacturing & Service Operations Management, and INFORMS Journal on Computing, Lu is also the recipient of the 2024-25 W. P. Carey Dean's Early Career Research Award, recognizing outstanding scholarship by tenure-track faculty in their first six years.

As Lu continues to focus on his research into the evolving relationship between people and AI, he says being recognized by INFORMS reinforces his commitment to advancing this important work.

"I look forward to building closer collaborations with industry partners and practitioners to produce impactful and socially responsible research, particularly in the age of generative and agentic AI that is transforming how we work, think, communicate, evaluate, and make decisions," says Lu.