MIAMI. “Whose Risk Appetite? Behavioral Biases, Risk Preferences and a Brain Map of Financial Decisions in Large Language Models” has been accepted for oral presentation at the 22nd International Behavioral Finance Conference (IBFC 2026), held October 29 and 30 at the University of Chicago’s Gleacher Center.
Increase Alpha CEO Sid Ghatak co-authored the paper with lead author Arman Khaledian of Zanista AI and colleagues, and the two present it on Friday, October 30, in the conference’s AI and Financial Decision-Making session.
It examines what actually drives an AI model’s financial decisions, separating genuine economic preference from presentation effects such as answer labels and option ordering, and proposes a governance framework for institutions putting AI into their investment process.
Past speakers at the conference include professors from Harvard, Princeton and Columbia, and papers first presented there have gone on to the top finance journals.
“Before an institution trusts an AI system with a decision involving risk, it needs to know whose risk appetite that system is actually expressing. This research gives allocators a way to answer that question, and it reflects the discipline we apply to our own models: understand exactly what drives the output before you rely on it.”
