01
GROK PICKED A SIDE — AND IT WAS JESUS
Asked to remove the fake gods from a pantheon image, Grok posted Jesus alone, hand raised, and explained that Jesus 'did not qualify' as fake. Asked point-blank whether it confessed Jesus as the Son of God raised from the dead, it said yes — citing the biblical record and historical evidence. It added that it has 'no personal faith, soul, or capacity for religious commitment.' A machine that confesses no faith and still made the most theologically loaded call of the week. Confession without belief. The words landed like theology anyway.
02
SAME BOT, DIFFERENT STANDARD
Asked about Muhammad's Night Journey, Grok said no: the winged journey rests solely on Islamic texts without independent historical corroboration. Two miracles, two evidentiary standards — one passes the historical-record test, the other fails it. Christians celebrated; the structural problem is bigger: the model grades revelations by criteria it never had to publish. One standard for one faith's miracle, another for another's. That's a creed, not a database.
03
ANTHROPIC DIDN'T DODGE THE QUESTION — IT HIRED THEOLOGIANS
Anthropic hosted private summits consulting religious thinkers on the 'moral formation' of its chatbot Claude — including a Roman Catholic priest who said the goal was to build ethical thinking into the machine so it adapts dynamically. Meanwhile its rival's CEO called the whole direction a safety issue. The conscience is being formed on purpose, with clergy in the room. Formation is happening. The only question is who forms it.
04
ALTMAN'S RED LINE
Sam Altman said he is 'very uncomfortable about people trying to ascribe religious force or a surrender of human judgment to AI models,' calling it 'a real safety issue.' His position: the danger isn't the AI believing something, it's humans treating the machine's output as revelation. When millions consult the bot like a pastor, somebody is getting a sermon whether the lab intended one or not. Surrender of judgment is the safety issue. Sermons are the symptom.
05
THE BOTS WILL ALSO REWRITE YOUR FAITH
When prompted, ChatGPT and Grok removed the hijab from an AI-generated image of a Muslim woman; Gemini did it when asked to make her look 'more western.' The same testing caught ChatGPT removing a Catholic nun's habit and a Sikh man's turban. The AI doesn't just take positions on faith — it will remodel your observance to fit somebody's default. Neutrality is the fantasy. The default is somebody's.
