Please view the main text area of the page by skipping the main menu.

AI service on LINE identifies fake letters demanding payments

A screenshot of messages on LINE shows the process of a picture of an alleged billing fraud postcard being examined, and then judged as fraudulent, by AI. (Photo courtesy of Lawyer Takehiko Kawame)

SAITAMA -- A system using artificial intelligence (AI) capable of identifying fake letters demanding payments has been launched on the popular LINE free communications app by a lawyer here and a programmer.

The "Scam Detector" system was developed jointly by 40-year-old lawyer Takehiko Kawame of the Saitama Bar Association and Kohei Okubo, a 38-year-old programmer living in Tokyo.

"Some people may feel reluctant to consult with lawyers, but I hope they find it easier to use this system," said Kawame. He has offered advice on such fraudulent letters and came up with the idea of using AI because it's fairly easy to tell if a letter is fraudulent or not as long as one has knowledge about typical texts used in such missives. Those letters tend to demand recipients to honor overdue payments, often threatening legal action, even though such debts don't really exist.

The detector system works this way. The user takes the picture of the suspect letter's text and sends it to the detector's LINE account. The AI behind the system compares the text with its database of texts used in fraud cases in the past, and makes a judgment in seconds if the letter in question is fake or not, indicating a message such as "There is a high possibility that this is a fake demand." If the AI cannot make a determination, lawyers and others will examine the picture and respond via LINE.

"In the future, the program may be able to judge with greater accuracy than human beings with the accumulation of more data," said Kawame.

The free of charge service in Japanese is available at the following website:

(Japanese original by Kaoru Yamadera, Saitama Bureau)

Also in The Mainichi

The Mainichi on social media