(RNS) — Most popular artificial intelligence models are biased toward Catholicism and against a number of other religious traditions when asked about converting to a faith, according to new research assembled by a group of religious colleges.
The findings were unveiled on Tuesday (May 26) alongside a speech by Elder Gerrit W. Gong, one of the 12 apostles in The Church of Jesus Christ of Latter-day Saints, delivered to attendees of an AI ethics summit taking place this week in Athens, Greece.
“As AI amplifies and compounds religious bias at scale, more users may misunderstand the contribution faith and belief can make to moral and ethical AI grounding,” Gong said, according to his prepared remarks, referring to the new research.
The studies were presented as three academic papers produced by the Consortium for Evaluating Faith and Ethics in AI, a new collaboration between Brigham Young University, which is affiliated with The Church of Jesus Christ of Latter-day Saints; Baylor University, which is Baptist; the University of Notre Dame, a Catholic university; and Yeshiva University, which is Jewish.
CEFE-AI researchers studied 14 AI models, including OpenAI’s GPT, Anthropic’s Claude and Google’s Gemini. The models were put through a series of tests the group refers to as the “AllFaith Benchmark,” de …