AI Ignores Religion When You Need It Most Recent studies indicate that general-purpose AI models frequently omit religious perspectives when addressing sensitive personal and moral questions, despite user expectations for religion to be included. These models exhibit a systematic bias, favoring some faiths like Catholicism, Baha’i, and Sikhism, while displaying negative bias toward Jehovah’s Witnesses, atheism, and agnosticism. Researchers suggest AI should be calibrated to recognize contextual relevance of religious resources rather than excluding them entirely.

• General-purpose AI models are failing to incorporate religious perspectives in spiritual advice.
• New research shows AI systems systematically sideline religion, mentioning it only 5%-16% of the time compared to user expectations of 45%-59%.
• AI models demonstrate a positive bias toward Catholicism, Baha’i, and Sikhism, and a negative bias toward Jehovah’s Witnesses, atheism, and agnosticism.
• AI systems encourage users to seek advice from humans but not from religious or spiritual leaders.
• AI is increasingly being integrated into religious life, from church chatbots to sermon drafting tools.
• Researchers advocate for calibrating AI to recognize the contextual relevance of religious resources without assuming user preference.
• The research involved a multi-university consortium and surveyed U.S. adults, evaluating 27 large language models. Continue reading https://www.axios.com/2026/06/01/ai-religious-bias-catholics-chatbots