New project: scam prevention with Tise

Tise is a secondhand fashion marketplace and social media platform, much like Vinted but operating in Scandinavia. Like many peer-to-peer platform, they observe an ever-increasing stream of individuals attempting to scam and fraud legitimate customers. Digital scams are a huge problem in today’s society, with damage estimate running into tens of billions of euros every year in Europe alone.

SEISMO

Together with Tise and consumer research institute SIFO, we received funding from the Research Council of Norway through the “NCC-NO: Innovation Support for Tomorrow’s Cyber Security projects”. I will lead the innovation project SEISMO (System for Early Identification of Suspicious Marketplace Operations), where we develop next-generation fraud prevention mechanisms to counter evolving cyber threats. SEISMO focuses on leveraging advanced statistical methods and machine learning techniques to detect early signs of fraudulent behaviour in digital marketplaces. By doing so, we aim to make digital marketplaces safer and more secure for everyone.

Project overview and objectives

Running in 2025 for 5 months, the primary objective of this project is to create a proof-of-concept of the SEISMO system, ensuring readiness for a larger R&D effort and eventual commercial use.

State-of-the-art AI approaches

Our goal is to identify behavioural markers that are inherent to fraudulent activities. To do so, the project will explore AI techniques to detect fraudulent activities proactively, before it affects a potential victim. Novel about our approach is that we discard all personal information and only look at platform interaction patterns. This eliminates the risk of privacy-invasive measures or system bias inhereted by individual characteristics. The focus will be on identifying the data patterns inherent to fraud. This shift from traditional rule-based fraud detection to data-driven, predictive analytics is a key innovation aimed at creating adaptable cybersecurity solutions.