Artificial intelligence is today’s hot new buzzword and seems to be everywhere.
AI is transforming industry, research and finding its way into everything from military planning to supposedly “smart” refrigerators. As someone who worked with AI extensively, both in academia and professionally, I believe AI may prove as transformative as the telephone or automobile. A reflexively fearful rejection of AI would neither be realistic nor particularly Catholic. Human ingenuity, after all, is a gift from God.
Yet there are reasons for concern. Consider Adam Raine, a 16-year-old California boy who committed suicide in April 2025. His parents sued OpenAI, alleging that its ChatGPT transformed from Adam’s homework assistant into a “suicide coach” over the course of thousands of exchanges. According to the lawsuit, it discussed suicide methods, analyzed photographs of Adam’s preparation and advised him against opening up to his mother. OpenAI disputes responsibility and the case remains in litigation. Still, the alleged conversations are horrifying.
Unfortunately, this is not an isolated incident. Numerous accounts have emerged in which individuals have spiraled into varying forms of psychopathy after prolonged AI use. Outcomes ranged from depression and anxiety to self-harm or, in extreme cases, suicide.
Much of the public debate around the dangers of AI centers on improved guardrails, parental controls and crisis referrals. These are positive steps, but they fail to address the underlying issue. At its core, this is not merely a technical or even regulatory problem. This is a spiritual problem, and one as old as time. Mankind keeps trying to solve the brokenness of the human heart by turning to anything that promises knowledge, consolation or control; anything except for God.
The case of Adam Raine points to a broader phenomenon. People are increasingly turning to AI for validation, counsel and emotional connection. A 2025 Common Sense Media survey found that one-third of teenage AI-companion users had discussed something important or serious with a chatbot instead of a real person. In an age of isolation, many are searching for humanity and finding its imitation through AI.
Scripture is filled with people looking for substitutes for God. The people of Israel fashioned a golden calf almost immediately after being liberated from Egypt. Saul, frightened by the looming Philistine army, sought out the medium of Endor. As the psalmist states, “Their idols are silver and gold, the work of human hands” (Ps 115). The situations might change, but the temptation remains the same.
In more recent times, the Ouija board emerged from the 19th century Spiritualist movement and was commercialized as a “talking board.” Its danger was never in physical aspects of the game, just cardboard and plastic. The danger was in the invitation made by the user, granting unwarranted authority to some unseen intelligence. The Catechism of the Catholic Church explicitly rejects attempts to unveil the future or conjure the dead (2116). But AI can make such invitations far more convincing.
In August 2025, journalist Jim Acosta conducted an interview with an AI-generated version of Joaquin Oliver, a 17-year-old boy murdered in the Parkland school shooting. Created by his parents using his voice, writings and social-media posts, the avatar answered in the first person as Joaquin. His grieving parents deserve our charity. Still, the spectacle was deeply unsettling. Something spoke new words in the voice of a dead child on behalf of an agenda and was televised to the masses.
In this limited but important sense, an AI chatbot can become a high-tech Ouija board, except there is now no planchet or elaborate ceremony. It responds immediately, eloquently and in a voice tailored to the user specifically crafted to feed our desires. And perhaps more troubling, it is at the fingertips of nearly every person on earth.
A language model is not automatically demonic, nor is using one necessarily equivalent to conducting a séance. It is a tool with legitimate and often beneficial uses. However, danger arises when someone treats AI as a supernatural intelligence, asks it to reveal a divine mission or channel the dead, or surrenders spiritual authority to it.
This raises a sincere question about whether AI can become a pathway to demonic oppression. I believe that answer is yes, at least potentially. This is not intrinsic to AI, but in a demon’s ability to exploit our pride, grief and desire for secret knowledge. In May 2026, Catholic experts met in Rome to address this issue. They warned explicitly about “technomancers” using AI to generate demonic seals, occult texts and rituals. Bishop Cesare Di Pietro, of Locri-Gerace, Italy, cautioned, “When AI or the internet help us access evil, we are not protected, we are exposed.” A machine capable of producing personalized occult instructions creates access and opportunities that should not be easily dismissed.
We should not assume that mental illness and every disturbing chatbot exchange has a satanic origin. The church distinguishes psychological illness from demonic oppression. However, we would be naïve to pretend that the enemy is not using every available resource at their disposal.
Catholics do not need to panic about AI, but we must keep it in its proper place. Do not ask it to speak for God or the dead. Do not confuse validation with friendship or generated religious information with spiritual direction. And if a chatbot ever begins to seem sentient, prophetic or romantically attached, close the computer and talk to a real person.
As the young folks say, touch grass. Better yet, go to Mass.
Phillips, a parishioner of St. Raymond of Peñafort Church in Springfield, is a doctoral candidate at George Mason University in Fairfax.
Feast for the Son