In the sunlit kitchens of Greece, the tradition of brewing Greek coffee has always carried with it more than just warmth and aroma. For centuries, women have peered into the remnants of coffee grounds left in small ceramic cups, seeking signs of fate, fortune, and future. But no one ever expected that in 2025, a language model trained to generate text would become the unlikeliest fortune-teller—and home-wrecker—of them all.
The tale of a Greek woman filing for divorce after 12 years of marriage because ChatGPT allegedly spotted signs of infidelity in her husband’s coffee cup is as bizarre as it is revealing. According to reports, the woman, fascinated by the growing trend of AI-assisted tasseography, photographed the leftover coffee grounds in their cups and asked ChatGPT for an interpretation.
The chatbot reportedly “saw” signs of betrayal. It claimed her husband was fantasizing about a woman whose name started with the letter “E” and even suggested that this mysterious figure was trying to tear their family apart.
What began as a quirky trend ended in a legal battle. When the woman informed her husband of the divorce, he was blindsided. According to an interview he gave on the Greek morning show To Proino, she didn’t even discuss it with him.
“She made us coffee, took some pictures, and uploaded them to ChatGPT,” he said. “I thought she was just playing around. But she wasn’t. She asked me to leave and told our kids we were getting divorced. A few days later, her lawyer called me.”
Blending Tradition with Technology: A Dangerous Brew
At first glance, the story feels almost laughable—another example of how far people can go when they take technology too seriously. But at the heart of it lies a complex interplay between ancient beliefs, modern AI, and personal emotions. Tasseography is a practice passed down through generations.
It’s not just about the patterns of coffee sludge. It involves an understanding of foam formations, swirling motions, cup rotation, and even the symbols formed in the saucer. Interpreting these signs requires experience, intuition, and often a personal connection between the reader and the querent.
This line ChatGPT told me that it replied to that woman when she uploaded images of coffee cups:
“There are signs of emotional unrest in this cup. A presence of someone whose name begins with the letter ‘E’ is strong. This person may be interfering in your relationship. The swirls indicate secrecy, and the cluster near the handle suggests emotional distance between you and your partner.”
But in this case, the woman bypassed tradition in favor of automation. Instead of consulting a human expert, she let a machine trained to predict the next word in a sentence analyze images of coffee grounds. What she received wasn’t prophecy. It was prediction—based on patterns in language, not in coffee. And yet, it was enough to shatter her faith in her marriage.
Read : These Are the Top Ten Most Famous Destinations Suggested by ChatGPT Around the World
The chatbot doesn’t have eyes. It doesn’t understand betrayal, marriage, or emotions. What it produces are echoes of the data it has been trained on: novels, dramas, love stories, confessions, gossip columns, and psychological profiles. Ask it to interpret something esoteric like coffee grounds, and it will oblige by spinning a story—because that’s what it’s built to do.
Fear, Fantasy, and the Echo Chamber of AI
So why did this woman believe ChatGPT over her husband? The answer might lie not in the AI itself but in the mindset of the user. According to her husband, she had a history of leaning into alternative belief systems.
“For almost a year, she was taking advice from an astrologer,” he said. “Now this.” Her journey through spiritual and mystical guidance systems had simply taken a digital detour. But that detour had very real consequences.
Experts in human-computer interaction have warned about the phenomenon known as “cognitive projection”—where users begin to imbue AI tools with authority and wisdom far beyond their actual capabilities.
Read : ChatGPT Diagnosed 27-Year-Old Marly Garnreiter with Blood Cancer a Year Before Doctors Found It
When someone is already harboring doubts, suspicions, or fears, an AI’s speculative or imaginative answer can act as validation. It’s not that the machine is speaking truth; it’s that the user is looking for something—anything—to justify what they already feel.
In this case, the suggestion that a woman with a name starting with “E” was undermining the marriage was a fictional flourish—one that may have confirmed the wife’s inner suspicions. The chatbot’s words were likely the result of random association based on linguistic training, not any sort of divine or digital insight. But they were taken as gospel.
The fallout from this incident has sparked a debate across Greek social media platforms and news outlets. Professional tasseography readers have weighed in, criticizing the process and its legitimacy. One reader explained, “You cannot simply take a photo and expect an algorithm to tell you the truth. It’s about energy, tradition, and trust. This is not tasseography. This is AI hallucination.”
When Algorithms Become Oracles: A Cautionary Tale
The Greek divorce drama is more than tabloid fodder. It’s a snapshot of a rapidly changing world, where technology no longer just assists us—it increasingly shapes our perceptions, decisions, and relationships. As AI grows more sophisticated and accessible, it’s becoming a tool not just for productivity but for introspection, therapy, creativity—and now, divination.
This new trend of AI-assisted fortune telling isn’t isolated. Around the world, people are turning to chatbots for tarot readings, astrological forecasts, dream interpretations, and yes, tasseography. Social media platforms like TikTok and Instagram are full of users sharing AI-powered spiritual revelations. Some see it as harmless fun. Others treat it with deadly seriousness.
But there are limits—dangerous ones. AI lacks context, conscience, and empathy. It doesn’t know when it’s crossing a line or suggesting something potentially harmful. The Greek woman’s case is a perfect example. A machine created to simulate conversation inadvertently became the catalyst for the breakdown of a family.
Her husband’s legal team is fighting the divorce, arguing that AI-generated suspicions should not hold weight in a courtroom. “He is innocent until proven otherwise,” said his lawyer. Whether or not the courts will agree remains to be seen. But even if the legal system sees through the coffee grounds, the emotional damage may already be irreversible.
This case also raises broader questions: How much trust should we place in artificial intelligence? Should AI have any say, even indirectly, in matters of personal life, love, and law? And most importantly, what happens when people begin to believe that machines know their hearts better than they know themselves?
In the end, this isn’t really a story about AI at all. It’s a story about belief. About how humans, in their eternal quest for answers, often find themselves turning to unlikely sources. Whether it’s the stars, the swirl of coffee sludge, or a chatbot trained on internet text, the need to make sense of life’s mysteries persists.
But when belief overrides logic, when speculation becomes certainty, and when a string of computer-generated words can dissolve a dozen years of shared life, we must ask ourselves: Is it really the machine we’re listening to—or just the echoes of our own doubt?
Because ChatGPT didn’t destroy this marriage. The coffee grounds didn’t whisper secrets. The real culprit may be something much older than any technology: the human tendency to see what we want to see, and to believe what we already fear.