.svg)
Dr. Joseph Breeden
Chief Executive Officer
CALM XAI
Dr. Breeden founded CALM XAI (Concept-Aligned Language Models for eXplainable AI) in 2025 to solve the problem of explainability and governance in language models. The CALM approach replaces guardrails and stops jailbreaks by compiling governance rules as direct overlays within the LLM's reasoning pipeline. This is made possible by a dramatically new approach to tokenization and conceptual alignment throughout. This has led to 15 patent filings and multiple research papers, so far.
Dr. Breeden has founded or co-founded Prediction Company (1991), Strategic Analytics (1999), and Deep Future Analytics (2011). He is a member of the board of directors of Upgrade, a San Francisco-based FinTech; an Associate Editor for AI and Ethics, the Journal of Credit Risk, the Journal of Risk Model Validation; and President of the Model Risk Managers’ International Association (mrmia.org).
Dr. Breeden invented vintage analytics for lending in 1997 and created credit risk models through multiple global financial crises. In 2018 Dr. Breeden invented Multihorizon Survival modeling, combining vintage analytics with behavior scoring using logistic regression or machine learning.
Dr. Breeden earned a Ph.D. in physics, and has published over 100 academic articles, 23 patents, and 8 books.
Speaking at SFF 2026
-
3:00–3:30pmThu 19 NovPolicy StageHall 3, Singapore EXPOOpen
Policy Rewiring
Digital Autonomy in AI: Optionality Over Absolute Independence
-
Prof. Dr. Fritzi Köhler-GeibMember of the Executive BoardDeutsche Bundesbank
-
Dr. Joseph BreedenChief Executive OfficerCALM XAI
-
Steven MaijoorExecutive Board Member & Chair of SupervisionDe Nederlandsche Bank N.V.
-
Sunayna TutejaFormer Chief Innovation OfficerFederal Reserve SystemModerator
About this session
No bank can or should build every AI capability itself. The real strategic question isn't build versus buy; it's whether institutions retain meaningful choice over the technology they depend on. As agentic AI begins making decisions that affect customers, capital, and risk, financial institutions must ask: How much digital dependency is acceptable? Can they trust and explain the decision of models they didn't build? What happens when the AI underpinning critical operations is controlled by a single provider?
The panel explores what digital autonomy should look like in the AI era. Rather than arguing for complete self-sufficiency, the discussion focuses on optionality: the ability to choose between providers, maintain control over sensitive data, satisfy regulatory expectations for explainability, and preserve resilient fallback options when business or technology priorities change. The conversation will examine how banks can adopt AI at scale without sacrificing transparency, governance or strategic independence - and why genuine optionality may become one of the industry's most important competitive advantages. -
