AI Fortune-Telling Booths in Seoul: How Korean Divination Became an On-Demand Service
In a Seoul café or shopping district, a customer can step into a fortune-telling booth, enter a name and birth date, talk with an AI character on a screen and receive a printed reading. Reports in spring 2026 documented these unmanned services. They compress the entry point to a traditional consultation into a few minutes and bring “AI fortune-telling” from the phone screen into a physical retail space.
Low prices, short waits and a game-like process help explain why passersby try them. Those features do not tell us whether a result is reliable, how personal information is stored or whether the customer understands what they are buying.
How the in-person service works
AFPBB News reported on April 14 that Seoul cafés had introduced AI face-reading devices and fortune-telling booths. Customers entered personal information, selected a topic and received an answer in a conversational format; some services printed a card. The reported prices included 7,000 won for AI fortune-telling and 8,000 won for face reading.
A March 4 report by the Korea JoongAng Daily described Seoul venues combining speech recognition, generative chat and printed talismans. These are on-site reports about particular businesses and services; they do not establish that every fortune-telling venue in South Korea uses the same system.
Automation changes the entry point
An unmanned booth offers an immediate experience without an appointment or the need to share every private concern with a consultant. For a curious visitor, it may feel like interactive entertainment. A user looking for a detailed reading may find that a fixed process and limited input cannot account for a complex situation.
A human consultant may offer follow-up questions and listening; a booth may offer lower prices and easy access. The experiences can coexist. Comparisons should focus on depth of interaction, price, knowledge sources and data handling rather than announcing that machines have replaced people.
Birth details and face data need transparent handling
Saju readings often require a date and time of birth. Face-reading services may collect images of a person’s face. Before collecting information, an operator should explain what is retained, for how long, whether it is used to improve a model and whether it can be deleted. Pricing and privacy information should be visible before a customer types anything into an in-store device.
Generative AI can turn inputs into a fluent answer, but users should be able to tell what the system actually calculated, which rules it used and what was generated as natural-language text. If a device cannot explain its interpretive basis, the experience is better treated as entertainment than as authoritative advice for a major decision.
Mystilink’s product difference should be verifiable
These booths show the commercial appeal of a low-barrier experience. If Mystilink emphasizes an ongoing AI Agent conversation or support for several traditions, the product should show its inputs, calculations, knowledge sources and memory settings. Features that are not live should be labeled as plans, not compared vaguely with other vendors’ devices.
Convenience does not guarantee understanding
AI fortune booths make the experience easy to try. Product quality also depends on whether users understand how an answer was formed, where their information goes and how to use the result. Making those details visible builds more trust than packaging a machine as a “more objective master.”