The Jesuit AI Ethics Framework

Abstract

A framework for evaluating and designing AI applications, grounded in Catholic social teaching and Ignatian spirituality and offered to anyone building or adopting AI systems, in companies, nonprofits, public agencies, health systems, and universities. The framework draws on three layers and does not conflate them. Magisterial social doctrine supplies the content, the four permanent principles of the Compendium of the Social Doctrine of the Church read alongside recent Church documents on artificial intelligence. Ignatian spirituality supplies the method, a discipline for choosing among competing goods under uncertainty. The Rome Call for AI Ethics serves as the operational bridge, giving the principles operational form. Every quotation from these sources is verified against the primary text with paragraph numbers, and the verification record is public. The document contains a five-question assessment rubric modeled on the Ignatian practice of a daily review, eleven criteria for assessing AI initiatives, requirements for conforming implementations, and the verification standard. A reference implementation, with a working assessment auditor and design workshop, is at https://jesuitaiethics.org.