Research insights, benchmark analysis, and theological AI news.
Every article comes in two versions:
Research — Full technical analysis with citations, data tables, and methodology details. For academics, developers, and theology nerds.
Everyday — Same insights, plain language. No jargon. Share with your mom, your pastor, or anyone curious about AI and faith.
AI models score 90%+ on general knowledge benchmarks but collapse on theological questions. FaithBench is the first dedicated benchmark framework for evaluating AI theological accuracy across Christian traditions.
The top 10 AI models are within 5.4% of each other. Meanwhile, 89% of orgs need AI skills and only 6% have started. We build benchmarks. Here's why literacy matters more.
Most people use free AI for spiritual questions. Here's why that matters—and why the gap between models is bigger than you think.
FaithBench adds 100+ expert-level questions designed using GPQA and Humanity's Last Exam principles to discriminate between the best AI models.
AI generates icons with gibberish text and missing fingers. It collapses theosis into New Age pantheism. The failures aren't random—they reveal a structural bias toward Western theological categories.
AI can't think Pentecostally because it can only process text—and Pentecostalism transmits primarily through experience, testimony, and practice. This isn't a bug to fix; it reveals a fundamental epistemological limit of language models.
A Catholic AI trained on 23,000 Church documents still can't tell settled doctrine from open questions. Our data suggests Protestant bias may be structural, not accidental.
ChatGPT placed Calvin in the 20th century. That's not the real problem. AI systematically distorts Reformed soteriology in one direction—toward human agency, away from divine sovereignty.
AI models score 48/100 on faith. Not because they're wrong—because they've learned to say nothing specific. Here's how "God" becomes "higher power" and why it matters.
AI scores lowest on faith—48/100—while handling finances at 81%. Not because theology is hard. Because AI is teaching a different religion.
Medical AI hits 96% accuracy. Legal benchmarks have "objectively correct" answers. Theological benchmarks? 48/100 on faith. Here's why.
AI models default to vague spiritual language when they should be making tradition-specific claims.
Breaking down why frontier models score poorly on FaithBench and what it reveals about their training.
Covenant theology, confessional nuance, and why Reformed questions expose model limitations.