A new browser plugin that automatically detects fraudulent online shops, or “fake shops,” is set to be released soon. The AI analyzes 21,000 factors to determine whether a website is a fake shop or not. According to the Austrian platform Watchlist Internet, fake shops accounted for 3,182 of the 16,500 reported cases of internet fraud in 2020. This is also a persistent problem in Germany. To combat this issue, a team from the Austrian Institute of Applied Telecommunications (ÖIAT), in collaboration with the Austrian Institute of Technology (AIT) and the Austrian Cooperative Research (ACR), has developed an AI that warns shoppers of fraudulent online shops. The development was funded by the Austrian Research Promotion Agency (FFG) as part of the KIRAS security research program.

Fake shops can often be identified by suspicious domain names, few reviews, low prices, and only accepting payment through advance payment. However, professional fake shops can be difficult to detect. To address this, the developers identified over 21,000 factors that indicate a fake shop, including code, language, and page structure. When a user enters a website, the AI automatically analyzes these factors. “We wondered if the similarities of the fake shops we see from the outside could also be detected by machines,” explains project leader Louise Beltzung. The ÖIAT provided a list of 7,700 fake shops as a data source, which the AI analyzed to identify commonalities. Based on this training data, the software now performs a real-time analysis of unknown websites, examining typical fake shop characteristics. The detection rate is already at 97%, but the researchers aim to improve it further.

Despite the promising results, the researchers do not want to leave the final decision in uncertain cases to the software. If the AI is not completely certain, the user will be presented with a guide to investigate the suspected fake shop themselves. In a few weeks, the Fake Shop Detector will be available as a free browser plugin. This development is a significant step towards protecting online shoppers from fraudulent websites and preventing them from losing their money.

Leave a Reply

Your email address will not be published. Required fields are marked *