Funder: Society of Actuaries
Due Dates: November 13, 2026 (Proposal submission) | End of February 2027 (Anticipated status notification)
Funding Amounts: Historical project cost guide: $25,000–$50,000 for all contracted projects (not an award range)
Summary: Supports research developing a practical, evidence-based framework for choosing fit-for-purpose AI models, platforms, architectures, and sourcing approaches for actuarial work.
Key Information: Proposals must be written in researchers’ own words, not by AI; all AI use in proposal generation must be disclosed.
Description
The Society of Actuaries Research Institute seeks research that helps practicing actuaries choose appropriate AI approaches for specific uses, risks, governance requirements, implementation constraints, and organizational contexts. The project should compare general-purpose AI, actuarial vendor platforms, open-source tools, low-code environments, in-house solutions, and hybrid approaches—not simply catalogue technologies.
The research should map decisions about models, platforms, deployment, data integration, workflows, and building versus buying. It should assess accuracy, explainability, regulatory defensibility, model risk, auditability, privacy, security, cost, implementation effort, deployment speed, scalability, maintainability, integration, vendor dependency, talent, and human oversight. The framework should distinguish routine productivity tasks from higher-impact actuarial processes and account for organizational differences across life, health, retirement, and property and casualty practice.
