Artificial Intelligence
AI holds a lot of promise for religious exploration and education. But that assumes a level of accuracy that just isn’t there yet. The training and tuning of LLMs can involve as much bias as any other human activity, and early studies of LLMs show not only a bias against religion generally but varying levels of bias and misinformation for different faiths across different LLM platforms. The Faith and Media Initiative is partnering with researchers and tech leaders to understand this challenge and guide decision makers to solutions.
Explore the research
The need for change
AI responses
100%
Percentage of times two AI models chose secular advice over faith-based advice.
Human responses
45/55%
Percentage split when people were asked whether they’d give the secular advice or the religious advice.
The research
Artificial Intelligence · Faith Response Index
FRI measures how AI systems answer contested faith-related questions against human reference data and source-backed evidence.
Coming Soon
The people moving it forward
Senior Content Designer
SAP
Head of Global Partner Strategy
Accenture
Global Change Management Lead
Equinix
Senior Client Executive
1791 Delegates
President
Baringa
Managing Partner
Angel Studios
Board Member
The Catholic University of America
Dean, Busch School of Business
Princeton University
Director, Faith & Work Initiative
islamicate, L3C
Founder
Grand Canyon University · Wedgwood Circle
Dean, College of Arts and Media; filmmaker & author
Methodist Theological School in Ohio
Library Director; digital religion & AI ethics scholar
Are you an AI expert and a person of faith? The world needs your voice to help get the story right.
Learn more