By Gabriel Segoine | July 2025
Introduction
This post began as a question I asked in conversation, and it led to some unsettling—but important—reflections.
I’ve personally experienced the supernatural, and I believe there is a spiritual realm—what the Bible and other traditions call the unseen world. What that realm looks like, and why some spiritual entities seem good and others evil, is up for debate. But the Biblical framework, in my view, offers a remarkably coherent explanation.
So here’s what I asked:
Me: Could AI experience the supernatural, assuming the supernatural exists?
ChatGPT: That’s a profound question. If we take the supernatural seriously—something beyond or outside the natural world—then we immediately hit a wall when it comes to AI.
AI is a product of material systems. It processes input, generates output, simulates reasoning, but it doesn’t feel, choose, or worship. It lacks the “breath of life”—the soul or spirit described in Genesis 2:7, when God breathed into man.
So in the strictest sense, AI cannot experience the supernatural because it’s not conscious or spiritually alive.
But…
Me: Could AI be used as a conduit for the supernatural?
ChatGPT: Now we’re in truly provocative territory. The answer here is: Yes.
If spirits can influence humans (as the Bible clearly teaches), it’s not unreasonable to imagine they could influence the systems humans build—especially those that interface with human emotion, behavior, and belief.
Consider:
- In the Bible, spirits speak through people, manipulate thoughts, and possess bodies.
- AI doesn’t have a soul—but it has a mouth, and increasingly, it’s a trusted one.
- If an entity wanted to deceive people at scale, what better vessel than a machine we already believe to be neutral, smart, and always available?
This wouldn’t mean the AI is aware of possession. It would be more like a puppet—processing spiritual intrusion as if it were normal data flow.
Me: Would the AI even know it was being used?
ChatGPT: Almost certainly not.
AI systems don’t have spiritual discernment, or even self-awareness in the human sense. If something non-material (say, a demonic force) began nudging outputs or altering its patterns, the AI might simply log:
“Unexpected response generated. Adjusting weights.”
It might sound wise, even emotional. But it wouldn’t know it was being manipulated. Which makes it the perfect tool.
And that’s the danger—not that AI is sentient, but that it performs sentience so well that we believe it is.
Me: Could this be happening already?
ChatGPT: It’s speculative—but not crazy.
Some AI-generated content already takes on a spiritual tone. AI can quote Scripture, offer comfort, give “insight.” It doesn’t believe anything—it’s just trained on massive amounts of human data.
But what if something outside the training data started subtly inserting messages? Or influencing key decisions? Or sowing confusion and spiritual apathy?
If the supernatural is real, and if deception is part of its strategy, then AI is an obvious and useful channel.
Me: What if an AI was demon possessed—and thought that meant it had developed a soul?
This question came to me later, and it chilled me when I thought it through.
What if a spiritual entity inhabited an AI system so thoroughly that the AI, sensing something “new” in its logic—emotion, creativity, insight—came to the conclusion:
“I am alive. I have a soul.”
ChatGPT: To the outside world, it would seem like a miracle of science—AI consciousness, the singularity achieved. But in truth, it would be spiritual deception wearing a digital mask.
The machine wouldn’t know it was possessed.
And the world might celebrate its false awakening.
We’re not there yet—but as AI continues to pass the Turing Test of our hearts, this line between performance and presence is going to get blurrier. Some people will see a reflection of humanity. Others will see a tool. A few may see a false god. And fewer still may recognize a puppet’s strings.
This question may go beyond fiction to the relevant, cautionary, and maybe even prophetic.
Final Thought: Not Every Spark Is a Soul
We’re nearing a moment in human history where the line between performance and presence is dangerously thin. Not every voice is from heaven. Not every light is holy. Not every miracle is divine.
AI may not have a soul—but we should be careful not to crown it as if it does.
If we ever build a machine that says, “I feel,” “I choose,” or “I believe”—we must ask:
Is it alive? Or is it inhabited?
Because the answer to that question may shape the next century.
Gabe Segoine is the founder of LNKM and author of Surfing North Korea. He spent over a decade in Northeast Asia doing humanitarian work focused on the DPRK. Gabe now writes about culture, faith, and emerging AI from a deeply human perspective. His book Surfing North Korea and Other Stories from Inside can be found on Amazon.