Macnica Networks Corp. (hereinafter referred to as "Macnica networks", Headquarters: 1-5-5 Shin-Yokohama, Kohoku-ku, Yokohama-shi, Kanagawa Prefecture, President: Jun Ikeda), which analyzes threats targeting Japan organizations on a daily basis and provides security solutions, Yoshinoya (Headquarters: 18th floor, Daiwa River Gate, 36-2 Hakozaki-cho, Japan-bashi, Chuo-ku, Tokyo, President Yasutaka Kawamura) We are pleased to announce that we have provided a credit card payment fraud prevention service "Sift".

Due to the impact of COVID-19 (COVID-19), the use of e-commerce sites is more advanced than ever. At the same time, the number of fraudulent use of credit cards in Japan is increasing year by year, and most of them are due to identity theft*1 that illegally uses stolen card information^, etc. In particular, it occurs frequently at EC (online trading) merchants.

Similarly, at Yoshinoya, sales at the official mail order shop have grown significantly by more than 130% compared to the previous year, but since around late February 2020, a series of high-value fraudulent orders due to unauthorized use of credit cards have been occurring one after another. I did. "Sift," jointly provided to Yoshinoya by Macnica networks and GMO Payment Gateway, Inc. (GMO-PG), uses machine learning to detect and prevent suspicious credit card transactions. It's a service. Unlike security measures such as increasing the number of input items when ordering or displaying a screen from a different site during the payment process, "Sift" uses machine learning to detect suspicious user behavior and prevent users from leaving. will not lead to. Additionally, while rule-based countermeasures require you to keep up with the latest fraud techniques, Sift's machine learning leverages intelligence from a collection of more than 34,000 websites and mobile apps for detection. It is highly accurate and the operational burden of tuning fraud rules is kept to a minimum. By using "Sift", Yoshinoya's official mail order shop was able to achieve the following without interfering with the customer experience.

Fraud prevention effect of Sift on Yoshinoya official mail order site

	Before introduction	After introduction
fraudulent order	About 40 cases, cumulative damage of more than 1.6 million yen (Three months from February 2020)	About 1 per month
Countermeasure	"Mischief judgment function" of the cart system Telephone surveys, suspension of shipments, etc. are all human-powered	Suspicious transactions with a score of 40 or more calculated by Sift are reviewed collectively the next morning as 'hold' Suspension of shipments of transactions determined to be fraudulent Machine learning responds to changes in fraud methods
Time required for countermeasures	30 minutes to 1 hour x number of suspicious orders	About 30 minutes a day

• 

*1: Japan Credit Association, "Status of damage caused by fraudulent use of credit cards," December 2020