Equitable AI in Finance & Health: Sudip Gupta, PhD

“Knowing that my initial insights were on the right path gives me pride.”

Dr. Sudip Gupta, Professor of Practice, at Johns Hopkins Carey Business School, juggles many tasks as a professor of finance, researcher, core faculty for the Hopkins Business of Health Initiative (HBHI), and mentor to students. He collaborates with researchers from medicine, public health, economics, and social sciences to investigate financial and operational issues in healthcare systems. These projects are focused on analyzing large-scale healthcare data, evaluating cost structures, and developing financial models that support improved access and patient outcomes.

But his passion is the integration of artificial intelligence (AI) and machine learning in financial systems. He was the director of a top MS in Quantitative Finance program, where he brough AI and data science in its curriculum. His research includes algorithmic decision-making, predictive modeling to enhance accuracy as well as credit access in credit scoring and risk assessment predictions, using large language models and agentic AI to enhance financial decision making etc. At Hopkins, he applies this data methodology from credit evaluation to also determine personalized health risk, which sometime may be analyzed through wearable health options like smartwatches.

“I wrote my first natural language processing (NLP) paper in 2013, on using NLP in finance. That was one of the first AI, it was not just word count-based NLP paper, but applications of topic modeling, which is a real machine learning model in finance,” he said. “At that time, it was a bit early. Its main contribution was emphasizing that context over words matter in financial documents, which was difficult for others to recognize at the time. Seven or eight years later, when it won the best paper award and validated that the work was important. Similarly, I ventured into reinforcement learning (RL) during my Ph.D. years, which has gained wide relevance today, especially with applications like combining RL with large language models.”

Finance, Health, and AI

Dr. Gupta’s background in AI research provided the foundation for his applications in finance and healthcare. His research focused on predictive modeling and natural language processing (NLP) techniques, including topic modeling and reinforcement learning approaches. These methods formed the basis for his work in predictive analytics and algorithmic decision-making in financial systems.

“Just like the credit risk, I want to see if someone could predict health risk using alternative data related to personalized medicine. For example, the smart watches collect a lot of personalized data. We can use this type of alternative data to predict personalized health risk,” Dr. Gupta explained. “This is very similar methodology like my prior work on predicting credit risk using alternative data for individuals who do not have enough credit history, so the methodology is generalizable to multiple areas.”

His work in AI evaluates predictive models used in financial decision-making, where he examines model accuracy and interpretability to determine best practices for deploying AI in regulated industries. This includes exploring methods for mitigating bias, ensuring transparency, and making sure AI is being ethical.

“One of our papers is about financial inclusion- a huge population that is financially invisible because they do not have enough credit history, and hence, do not get enough credit that they need. It need not be that they are not credit worthy, rather the lack of enough credit history could be various,” he explained. “In an emerging market like India, it could be that you are an unbanked customer, or as a millennial you do not have enough credit exposure yet. Harnessing alternative data like social connectivity and digital footprints with the help of AI, we try to come up with an alternative credit rating”.

“The World Bank and International Finance Corporation, for example are trying to implement similar credit rating methodology, primarily to give loans to women in Latin American countries. Other, financial institutions and fintech companies are trying to implement it in different places as well,” he said. “You also don’t want to be biased by race or gender. Many times, race and gender are not in the model, but other correlated variables like location or income may introduce bias. For example, if you belong to a certain group or population, like women, you may not have enough banking exposure because most of the loans were taken in the names of the head of the (male) household. So, the historical data to train for credit risk will reflect a selection bias in data where women did not enough credit history and hence did not have a higher-grade score. If you use that data to train an AI model to give loan you may (unfairly) select against the women to get a loan.” His research is to model all the nonlinear channels appropriately, debias the algorithm, and improve predictions. This approach applies to both for credit risk as well as health risk.

Global Engagement & Policy Impact

Dr. Gupta’s work extends to international contexts, with projects evaluating financial inclusion and the use of AI-driven financial tools in emerging markets in Asia, with partners in the Indian School of Business, the Indian Institute of Management–Bangalore, and the National University of Singapore. These studies assess the accessibility, reliability, and effectiveness of credit systems that incorporate alternative data and machine learning techniques. By analyzing the performance of such systems, his research provides insights for both local policy design and broader discussions on equitable access to financial services.

He has also collaborated with government and regulatory agencies, including the Competition Commission of India (CCI), the Securities and Exchange Board of India (SEBI), and the Reserve Bank of India (RBI). In these roles, he provides guidance on policy frameworks, the deployment of AI-based financial solutions, and methods for evaluating market impact. His work informs regulatory decisions on topics such as algorithmic fairness, risk management, and financial inclusion. He is also a fellow of the Centre for Responsible AI at the Indian Institute of Technology, Chennai. He has served as an expert in antitrust matters and in various financial class action litigations.

“I’m very passionate about a project that I’m currently working on, which is the financial toxicity of cancer care.”

One of Dr. Gupta’s current research areas is financial toxicity in cancer care. Financial toxicity, he explains, is the economic strain and stress that patients and families experience because of treatment costs and associated financial burden. He looks at the financial burdens associated with cancer treatment and investigates the drivers of these burdens to evaluate interventions designed to reduce patient stress without compromising care quality. By combining financial modeling with patient data, he can identify patterns of cost exposure and financial risk to inform healthcare policy and program design.

“My role is mostly from AI and the finance part. Cancer affects the individual’s well-being and health, but the financial impact and the burden it has on not only the patient, as well as the caregivers and the immediate families is huge. It’s not just only in emerging market, but also in more advanced economies like the US,” he said. “The broad question is, could we devise something which could finance motivated healthcare strategies which makes this painful journey easier for the group of patients as well as their caregivers and families?”

As Dr. Gupta researches the challenges surrounding financial toxicity in cancer care, he highlights a critical intersection of health and economics that demands attention. Looking ahead, the integration of AI and alternative data in both finance and healthcare presents a unique opportunity to create systems that are not only efficient but also equitable. By addressing the complex factors at play, Dr. Gupta’s work aims to redefine accessibility in these sectors, ultimately fostering a landscape where financial and health-related decisions are informed, fair, and supportive of all individuals.