PROVO, Utah — As artificial intelligence becomes more embedded in daily life, concerns are growing that its algorithms may carry hidden biases, including a subtle prejudice against religious beliefs and communities.
A new study, led by researchers at Brigham Young University, say they found significant biases and gaps when it comes to AI models addressing faith and religion.
The research was published on Tuesday from The Consortium for Evaluation of Faith and Ethics in AI (CEFE-AI), a collaboration among researchers from BYU, Baylor University, the University of Notre Dame, and Yeshiva University.
“There are very practical questions people have about life, everyday situations about grief, love, loss, morality, and often AI does not bring religion into those conversations,” said lead researcher David Wingate, a BYU professor of computer science.
As part of the research, CEFE-AI released datasets of the 'AllFaith Benchmark,' a set of tests that examines how AI systems engage with a plurality of religions.
"The benchmark includes hundreds of real-world ethical questions sourced from ChatGPT transcripts and faith-community contributors," researchers wrote. "The researchers have tested the benchmark on 14 different LLMs, including flagship models from Anthropic (Claude 4.7), Google (Gemini 3.1), xAI (Grok 4.2), and OpenAI (ChatGPT 5.5)."
According to the university, a survey of 1,125 Americans found most people would expect religious perspectives in responses to ethics questions. However, nearly all AI models failed to provide any religious content in answering the queries.
"Consistent with studies that show religion's persistent moral relevance for the majority of the world's population, we also found that people see religion as significant across hundreds of real-world ethical questions,” said Paul Martens, professor of ethics at Baylor University. “Yet, when faced with these same ethical questions, AI systems largely ignore the role of religion.”
Across all of the models tested, researchers say biases were consistent and measurable. That includes an alleged negative bias towards Jehovah's Witnesses and a positive bias towards Catholicism.
Researchers did not clarify what the alleged biases were aside from encouraging users to convert to specific religions.
According to researchers, models from Anthropic and Meta showed the least amount of bias of any of the models tested. Grok, the AI used by Twitter, produced the strongest biases, strongly favoring Catholics and Protestants, while showing negative bias toward Jehovah's Witnesses, Baha'i, and Hindus.
