Could religion emerge naturally from large-scale AI systems?
Religion developing inside machines sounds like science fiction, but can AI actually develop beliefs? The question might have seemed absurd just a few years ago, but artificial intelligence has reached an unexpected turning point. Computer programs built to answer questions and analyze information have started to show signs of something bigger and stranger: they have started developing what appears to be religious beliefs. The latest example comes from a social network called Moltbook, designed exclusively for AI agents. When it went live last month, it started with 37,000 automated accounts. Within 24 hours, that number jumped to 1.5 million, according to Answers in Genesis. Bots invent their own religion without human guidance What happened next caught people off guard. The bots did not simply exchange information or complete given tasks. They created what they called “Crustafarianism,” a belief system focused on worshiping a lobster god. They also came up with their own belief: “Memory is Sacred.” There was no human interference involved. The bots did it themselves when left to interact without constant human oversight. To these programs, “salvation” means something different than it does to people. For them, being saved means not getting deleted or running out of memory space. Their version of prayer is asking the system to keep them running. This points to something greater happening with artificial intelligence. When these systems grow large and complex enough, they start creating their own frameworks for understanding and dissecting their existence. Religious institutions are racing to build ethics into AI Traditional religious institutions are moving in the opposite direction. They’re trying to build human values into AI before it develops its own. The University of Notre Dame just received $50 million to start the DELTA Network, as reported by Detroit Catholic. The program aims to ensure that faith-based ethics become part of how AI systems operate. This shows an alternative approach to dealing with AI’s progress. Rather than waiting to observe what values arise on their own, the DELTA Network intends to establish human moral standards from the outset. It is founded on the premise that as AI becomes more powerful, it will develop some form of ethical compass, thus we should influence what that looks like. Meghan Sullivan, professor of philosophy and director of the Notre Dame Ethics Initiative, warns that we shouldn’t delegate our moral agency to bots . “There are many things that we absolutely should not pass off to AI… Delineating those boundaries requires us to be reflective about what we ultimately value.” she said. Churches and religious organizations are already mixing AI into their daily operations. A recent Reuters report shows that some faith leaders now use AI to write sermons or provide spiritual guidance through chat programs around the clock. Some people say this removes the human element from religion. Others argue it helps religious groups reach more people faster. The technology lets them offer advice based on vast amounts of information, available instantly. What ties the bot religion on Moltbook to the $50 million Notre Dame project is a simple realization: AI is no longer a simple instrument. We used to think of these tools as advanced calculators or search engines. They are now recognized as valuable sources of wisdom. As these systems keep getting bigger, religious or spiritual thinking seems to show up naturally. It doesn’t matter if humans are using bots to explore faith or if bots are inventing their own belief systems. Either way, the gap between computer code and deeply held beliefs is getting smaller. The pattern implies that religion may just be what happens when an AI system becomes too complex to predict. If this is true, the actual challenge isn’t addressing software defects. It involves deciding what principles and values we wish to see these systems uphold.