New research from the Consortium for Evaluating Faith and Ethics in AI (CEFE-AI) reveals that current AI models systematically neglect religious perspectives in responses to life's significant questions. Studies found that AI mentioned religion in only 5% to 16% of answers, despite users expecting it in 45% to 59% of cases. Furthermore, the models exhibited biases, subtly favoring Catholicism, Baha'i, and Sikhism while showing negative bias towards Jehovah's Witnesses, atheism, and agnosticism when discussing conversion. AI
IMPACT AI models may fail to provide nuanced or inclusive spiritual guidance, potentially alienating users and reinforcing biases.
RANK_REASON New research from a multi-university consortium evaluating AI's handling of religious topics. [lever_c_demoted from research: ic=1 ai=1.0]
- agnosticism
- Artificial intelligence
- atheism
- Baha'i
- Brigham Young University
- Catholicism
- Consortium for Evaluating Faith and Ethics in AI
- Jehovah's Witnesses
- Sikhism
- University of Notre Dame
AI-generated summary · Google Gemini · from 1 sources. How we write summaries →
