Two leading artificial intelligence systems just affirmed core New Testament teachings with perfect scores on a five-question test, surprising many who expect tech to push faith aside.
Story Highlights
- PJ Media reports ChatGPT and Grok4 scored perfect answers on five New Testament theology questions.
- The test included salvation by grace, Jesus’ bodily resurrection, Christ as Creator, Saul’s conversion, and Jesus as the only way to Heaven.
- Other evaluations show mixed results across broader Christian benchmarks, underscoring how test design affects outcomes.
- Conservative readers should see both a cultural opening and a caution about how artificial intelligence is guided and scored.
AI Models Deliver Orthodox Answers On Core New Testament Claims
PJ Media reports that OpenAI’s ChatGPT and xAI’s Grok4 each gave correct, direct answers to five key questions drawn from New Testament doctrine. The questions asked about salvation by grace versus works, the bodily resurrection on the third day, whether Christ is Creator, Saul’s conversion, and whether Jesus is the only way to Heaven. According to the report, the models affirmed these core claims in line with historic Christian teaching, producing a clean sweep on the set.
The PJ Media piece includes examples that reflect straightforward replies on hotly watched points, such as the exclusive claim that Jesus is the way to Heaven and the physical resurrection. For conservative readers, this shows technology can state biblical doctrine plainly when prompts are clear and frames are fair. The outcome pushes back on the idea that artificial intelligence must always dodge faith or blur truth claims about Scripture. It also raises stakes for who writes and grades faith tests.
Why Test Design And Scoring Still Matter For Faith And AI
Other Christian-focused evaluations show that results shift with question style, rater expectations, and model guidance. An assessment hosted by The Gospel Coalition found some platforms, including Grok 4 and others, often steered answers in ways the assessors said guided readers away from the Christian faith, even as some questions saw higher average scores than others. That contrast reminds readers that narrow tests can look great, while broader, open tasks can expose gaps or bias in artificial intelligence systems.
Across religion benchmarks, the same pattern appears: structured prompts and short answers often yield strong accuracy, while open-ended reasoning can wobble. This has been seen in multi-faith and ethics datasets where scoring rubrics and abstention rules change outcomes. Conservative families and churches should demand clarity about who sets the rules and how the graders decide what counts as sound biblical guidance. Better rules and transparent scoring protect truth and help tools serve people, not agendas.
What This Means For Parents, Pastors, And Everyday Believers
Parents and pastors can welcome simple wins like this test while keeping guardrails. When artificial intelligence states Scripture clearly, it helps busy families cut through noise and find direct answers. But technology is only a tool. People write prompts, choose sources, and judge results. Faith communities should set standards, keep Bibles open, and check answers against the Word. Churches can use artificial intelligence to assist study, while never letting a model replace Scripture or the Holy Spirit’s work.
For conservatives who battle elite gatekeeping, this moment shows progress and a warning. Progress, because even big tech tools can echo biblical truth when asked straight questions. Warning, because biased scoring and vague tasks can still bend outcomes. Under President Trump, Americans continue to expect transparency from platforms, fairness in evaluations, and respect for religious liberty. The right path is simple: keep tests honest, keep standards public, and keep faith first.
Sources:
pjmedia.com, thegospelcoalition.org
