Skip to content

How AI Is Reshaping the Spiritual-Tech Market: Personalisation, Privacy and Ethics in Astrology and Tarot Apps

ai generated photo of a person holding a phone

On May 4, 2023, the astrology app Co–Star added a feature called “Ask the Stars.” Users could type any question, from whether to get bangs to whether to quit a job and move abroad, and receive an answer built from their birth chart, the current positions of the planets and the company’s database of interpretations. The first three questions were free; after that, one more cost 99 cents.

It was a small product update with large implications. For most of the last century, astrology content was written once and read by millions: twelve signs, one paragraph each. Generative AI makes it possible to produce a different, fluent answer for every person and every question, on demand. By December 2023 The Atlantic was running a piece titled “AI Astrology Is Getting a Little Too Personal.”

Not every product in the category runs on a chatbot. Many sites still offer free online astrology tools built on fixed methods rather than generated text. AstriOnLine, for example, gathers a birth chart calculator, a tarot reading, a destiny matrix and a lunar calendar on one page, and states plainly that astrology, tarot and numerology “don’t describe an inevitable future and don’t replace personal decisions, professional advice or verifiable facts.”

Personalization: Calculation Versus Generation

The part that is just math

It helps to separate two very different kinds of “personal.” Computing where the planets were at the moment of your birth is astronomy. AstriOnLine’s birth chart page, for instance, says it converts local birth time to universal time using historical time-zone rules and computes positions from astronomical ephemerides, matching professional figures to within a few arc seconds. Enter the same data twice and you get the same chart.

The part that is generated

Interpretation is different. A large language model asked “why am I self-sabotaging?” will produce an answer that sounds intimate and authoritative, and it may produce a slightly different one tomorrow. Nothing about that is unique to astrology, but the setting matters. People often open these apps at uncertain moments, and a confident paragraph tailored to their chart can carry more weight than a generic horoscope ever did.

Different rules, different results

Even the deterministic side has a wrinkle users rarely see. Numerology and astrology contain competing schools, and an app’s output depends on which one it follows. AstriOnLine’s destiny matrix page notes that different methods can produce different maps or names, and its life path guide explains why another site may show a different number: some schools reduce the day, month and year separately, and master numbers such as 11, 22 and 33 are handled in different ways. A trustworthy tool says which rules it uses.

Why fixed text still has a place

Pre-written interpretations have an unglamorous advantage: they can be reviewed. An editor can check a dictionary entry for a planet, card or number once and stand behind it. A generated answer is new each time, which makes quality control far harder and puts more responsibility on the design of the system around it.

Privacy: Your Birth Data Is Personal Data

What these apps collect

A natal chart requires a date, an exact time and a place of birth. Combined with a name or an email address, that is a fairly precise identifier. Add the questions people type into a chatbot, about relationships, money, health or work, and an astrology app can end up holding some of the most revealing text a person writes all year.

A cautionary precedent

The wellness sector has been here before. In 2021 the U.S. Federal Trade Commission settled with Flo Health, maker of a period-tracking app, over charges that it had shared users’ sensitive health information with analytics and marketing services, including Facebook and Google, despite promising privacy. The settlement required, among other things, that Flo obtain users’ consent before sharing health data in future. Spiritual-tech apps handle a different kind of intimacy, but the lesson carries over.

What good practice looks like

Some design choices are simple. AstriOnLine’s life path calculator says the date entered “stays in your browser and is never stored,” and its tarot tool asks only for a name or a pseudonym, warning users not to enter health, financial or legal details in their question. Apple’s App Store privacy labels, required since December 2020, give users another quick check on what an app collects and whether it is linked to their identity. Reading them takes a minute.

Ethics: Drawing the Lines

Vulnerable moments

Most users treat these products lightly. A 2024 Pew Research Center survey found that only 1% of American adults rely a lot on astrology, tarot or fortune tellers when making major decisions, with another 5% relying a little. But 6% of a large user base is still a lot of people, and they are precisely the ones most exposed to persuasive, personalized language. The European Union’s AI Act already prohibits AI systems that use manipulative techniques or exploit vulnerabilities in ways that cause significant harm.

Honesty about the machine

Users deserve to know when an answer was written by software rather than a person, and what that software was built on. Clear labelling also protects the companies: an AI reading presented as a human consultation invites exactly the kind of backlash the sector can least afford.

Knowing when to hand off

The most responsible tools are explicit about their limits. A chatbot that receives a question about self-harm, a medical symptom or a legal problem should point to real help, not to a transit of Saturn. Writing those boundaries into the product, rather than burying them in terms of service, is a design decision, and a test of whether a company takes its users seriously.

Where the Market Goes Next

AI will keep making astrology and tarot apps more personal, more conversational and more habit-forming. The products that earn lasting trust are likely to be the ones that keep calculation transparent, collect as little as they need and stay humble about interpretation. Personalization is the selling point. Privacy and restraint are what keep people coming back.

For users, a short checklist goes a long way. Before installing an app, read its privacy label and find out whether your birth data and questions are stored, shared or used to train models. Use a pseudonym where the service allows it. Keep sensitive details out of the question box. And treat any reading, generated or written by hand, as a starting point for your own thinking rather than a decision made on your behalf.

Leave a Comment