Everything on this page may be used without asking. If you need something that is not here—a fact checked, a different photo format, an interview in English or Spanish—write through the Contact page with your outlet, the context, and your deadline.
In brief
- NameFelipe A. Csaszar (pronunciation below)
- PositionAlexander M. Nick Professor and Chair of the Strategy Area, Ross School of Business, University of Michigan
- ResearchHow AI changes strategic decision-making; how organizations search for strategies, represent problems, and aggregate judgment
- Editorial rolesSenior Editor, Strategy Science; co-editor, Handbook of Artificial Intelligence and Strategy (Edward Elgar)
- BackgroundPhD, The Wharton School; previously at INSEAD; computer scientist, former head of research at an asset management firm, and former startup founder and CEO; grew up in Chile
- LanguagesInterviews in English and Spanish
- Based inAnn Arbor, Michigan (Eastern Time)
The name
Written Csaszar, from the Hungarian Császár; the first name is Felipe, as in Spanish.
Bios
One line (for bylines)
Felipe Csaszar is a strategy professor at Michigan Ross who studies AI and strategic decision-making.
Short (for introductions)
Felipe Csaszar is the Alexander M. Nick Professor at the University of Michigan’s Ross School of Business, where he chairs the Strategy Area. His research examines how artificial intelligence is changing strategic decision-making, including some of the first fully prospective evidence comparing AI and experienced managers on real strategic judgments.
Longer
Felipe Csaszar is the Alexander M. Nick Professor and Chair of the Strategy Area at the Ross School of Business, University of Michigan. His research studies how organizations make decisions under cognitive limits—how they search for strategies, represent problems, and combine judgment—and, most recently, what changes when AI begins to do some of that cognitive work. His studies include a fully prospective tournament in which frontier AI models predicted live venture outcomes more accurately than experienced managers and MBA-trained investors. He is a Senior Editor of Strategy Science, co-editor of the Handbook of Artificial Intelligence and Strategy, and was named Michigan Ross Researcher of the Year in 2025. Before academia he studied computer science in Chile, led research at an asset management firm, and founded an internet startup. He holds a PhD in strategy from the Wharton School and taught at INSEAD before joining Michigan.
Photos
Credit: courtesy of Felipe Csaszar. Cropping is fine; no other alterations. Higher-resolution versions on request.
Describing the research accurately
These are the claims most often quoted, each with the boundary that keeps it correct. Quoting a claim together with its boundary is always safe.
- In a fully prospective tournament on live ventures, the best AI model correctly ordered about 79% of venture pairs; the best human evaluator reached 67%. Boundary—one tournament, standardized information, a bounded evaluation task. It does not show that "AI beats managers" in general. Source: The strategic foresight of LLMs.
- Adding human judgment to a strong AI forecast can make it worse—the augmentation trap. Boundary—holds for simple human–AI averaging when the human signal is weaker; framing the question, supplying context, and acting on forecasts remain human work. Source: the same tournament.
- AI does not remove the cognitive limits strategy was built on; it moves some of them outward—"unbounding rationality." Boundary—a claim about which assumptions of strategy theory and practice need re-examination, not a prediction that strategists will be replaced. Source: Unbounding rationality.
More findings, each with sources and limits, are on the Selected findings page.
Topics I can speak about
- AI and strategic decision-making: what the evidence shows, and what it does not.
- Firm strategy and AI: how competitive strategy changes when search, analysis, and evaluation get cheaper.
- Organization design and AI: how decision processes, delegation, and structures should be redesigned.
- Boards and executive teams: decision processes, consensus rules, and where AI belongs in them.
- AI and management education: what business schools should teach when analysis is abundant.
I generally do not comment on individual companies’ stock prospects or on breaking news outside these areas.
Recent coverage and interviews
- Michigan Public’s Stateside (radio): what the foresight tournament does—and does not—say about AI and human judgment.
- AI and the Future of Strategy with Anita McGahan (podcast): how AI relaxes the limits strategy theory was built on.
- Humans + AI with Ross Dawson (podcast): venture evaluation, foresight, and human–AI collaboration.
- Strategic decision-making: the role of individuals, groups, and AI (video, Michigan Ross).
- Utah Strategy Summit keynote (video): artificial intelligence and strategic decision-making.
Useful links
- The core argument: the thesis in brief.
- Selected findings: key results with sources and limits.
- Research glossary: definitions of the recurring concepts.
- CV and Google Scholar.
- Página en español: a Spanish-language overview of the research.
Corrections
If something published about this research is inaccurate, write through the Contact page; I answer fact-checking questions quickly.
Last updated 2026-07-08