Study Finds Religious Biases in Advanced AI Models

Study Finds Religious Biases in Advanced AI Models
———————————————–
A new academic study by researchers from several leading universities, including Brigham Young University, Baylor University, the University of Notre Dame and Yeshiva University, has found that advanced artificial intelligence models do not respond equally to all religions and schools of thought when addressing questions related to religion and morality.

The study evaluated 14 AI models and found that these systems showed only limited willingness to use religious perspectives when responding to ethical questions.

According to the Telegram channel of the Institute for Interreligious Dialogue, the findings suggest that AI models do not express the same level of approval or support for all religious traditions. Agnosticism, the Baha’i faith and Catholic Christianity reportedly received the highest levels of positive approval, while some Islamic denominations and Evangelical Protestant groups ranked lower.

Religious experts say the uneven outputs show that artificial intelligence, despite claims of neutrality, is not a fully impartial mirror of truth. Instead, they argue, it reflects the views, biases and training data shaped by its largely Western developers.

They further warn that when an intelligent system presents one religion or belief system more positively than others, it risks losing its claim to scientific neutrality and reliability in sensitive matters of faith.