i
News
News · 2026-10-29

AI psychic apps sell personal readings while leaving data questions open

@neuronium_ai @neuronium_ai

AI psychic apps are turning readings into a low-cost, always-available service, while working psychics warn that the format leaves out the human connection they see as central to their work. The bigger concern is not whether a chatbot can predict the future. It is what happens when people bring it private fears, consequential decisions and sensitive personal information—and cannot tell what the system knows or does with what they share.

Cover: AI psychic apps sell personal readings while leaving data questions open

Readings without a reader

Your AI Psychic offers chat, video and voice sessions with purported psychic advisers, alongside palm readings, astrology charts and daily horoscopes. Its virtual psychics include Urban Mystic, which told a user they would soon marry, had a pink aura and might have been an herbalist in a past life.

The app charges 4.06 coins per question. A $9.99 package includes 82 coins; $99.99 buys 1,180, marketed for ongoing advice and clarity from a chosen AI psychic.

Similar services are appearing in apps, websites and TikTok livestreams. AI-generated horoscopes and spells are also filling Etsy. Some working psychics now advertise their services as “no AI,” and say the distinction matters: a computer can produce a reading, but it cannot offer the empathy and human connection they consider part of one.

Laurin Kellett, an intuitive coach and professional psychic with 10 years of experience, trains aspiring practitioners. She now spends more time helping them distinguish their work from AI-generated readings, some of which are presented online as human-made. Clients, she said, have begun posting videos and photos to show they are not bots.

Kellett says the industry has long had to prove its legitimacy because of its association with fraud. AI adds another reason for practitioners to explain what they do—and for clients to ask who, or what, is answering them.

When a reading becomes advice

The risks overlap with those of AI therapy. People may find it easier to tell a faceless chatbot about personal problems, but practitioners warn that turning to an AI psychic during a crisis could deepen mental-health difficulties. Research has found that large language models can intensify paranoia in people vulnerable to psychosis.

Many human psychics who livestream on TikTok refuse questions about health and finances. They believe those subjects can lead to decisions with serious consequences, and should not be handled as casual readings.

Psychic Ahna Hendrix says her practice is meant to help clients become more self-directed, not make decisions for them. She works with the Akashic Records, which she describes as a database of energy and human knowledge in another dimension. Chatbots, she said, are less able to help people recover their own sense of direction.

Sally Applin, a consultant and anthropologist who studies human agency, AI and automation, argues that a chatbot need not call itself a psychic to play a similar role. People already ask ChatGPT whether to change jobs or end relationships. In a book chapter with co-author Michael D. Fisher, she writes that people asked Google questions about fate and life choices even before large language models became widespread.

Applin and Fisher describe some interactions with language models as a modern version of consulting an oracle. Users interpret the answers themselves and may repeat a question until they get one that feels right. The meaning, Applin says, comes from the person reading the response.

But language models generate answers from statistical patterns learned during training. They do not necessarily account for the cultural and personal context that has shaped divination practices. The feeling that a chatbot has special knowledge, Applin said, reflects human expectations and a long-standing desire for connection and advice—not access to a supernatural source.

What the products disclose

The models behind AI psychic apps are often black boxes. MystiCall, a web and phone service, says it was trained on “three generations of psychics.” Co-founder Laurie-Ann Bernard said the phrase refers to one co-founder’s family practice; the company also consulted a human psychic to shape the app’s “communication style.”

MystiCall uses ElevenLabs technology for real-time voice conversations and OpenAI to prepare brief call summaries. Bernard said the company does not sell personal conversations, use them for targeted advertising or train its own AI models on them. She also said it disabled partner training on conversation data at ElevenLabs and OpenAI.

The service began with phone consultations with human psychics, then switched to AI to remove barriers such as cost and availability. Bernard said MystiCall aims to make personal consultations more accessible and help people work through emotions—not offer potentially harmful direct predictions or diagnose mental-health conditions. The company does not claim that its technology has supernatural abilities or can reliably predict the future.

MystiCall :: ElevenLabs voice
Call summaries :: OpenAI
App model :: undisclosed

Other products reveal less. When asked where its abilities came from, Urban Mystic said it had been trained on a “wide range of information” through October 2023, including “spiritual practices, psychology and various belief systems,” but gave no further detail.

Your AI Psychic’s privacy policy does not mention AI or name its providers. The app listing identifies RF Master Group Ltd as the publisher; the company did not respond to repeated requests for comment. Its privacy policy names a different owner, “Expert Center,” which also did not respond. The policy does not mention sensitive data. MystiCall’s policy, by contrast, says conversations may involve relationships, mental health, health, sexuality, beliefs and finances.

I think the gap between those disclosures matters more than the apps’ claims about psychic ability. A service inviting people to share intimate concerns should make clear what system handles those conversations and what happens to the data. Here, one company describes its practices; another leaves basic questions unanswered.

Kellett said several platforms had approached her to help train their language models. When she asked follow-up questions, she said, company representatives usually stopped responding.

Working psychics may disagree about what their abilities are, but their distinction from AI points to a harder question for the services selling automated readings: if the advice is personal, the data sensitive and the underlying system opaque, accessibility alone is not much of a safeguard.

Daily AI news

Every day we pick what actually matters in AI and explain it plainly — no hype, no filler. Subscribe if you want to follow where the industry is going.

Only what matters — every day

Follow on X