A survey of French e-commerce businesses on online fraud experiences and AI effectiveness in fraud prevention.
Extract
[...] - Very important - Moderately important - Not very important Respondent Name Percentage Very important 14 82,4% Moderately important 3 17,6% Not very important TOTAL 17 100,0% Source : Based on the data from the Google Forms survey (September 2024) 2.2.2 Discussion 2.2.2.1 Sector of the company According to the data of the people surveyed work in the online retail sector, which highlights its importance among participants. Following are online financial services and the technology and software sector (respectively 17.6% and 23.5%). We then note that 5.9% of people offer services offer online services. We observe that the majority of the sample works in the e-commerce and technology sectors. This explains the tendency of these companies to move towards digitalization. 2.2.2.2 Company size (number of employees) The majority of companies have less than 30 employees. [...]
[...] Other (please specify) Section Legal Framework and Regulations Are you familiar with European regulations (for example, PSD2 and two-factor authentication) and French regulations regarding online fraud prevention? How do you think they affect businesses? Free response 10- French companies are sufficiently prepared to comply with online fraud prevention regulations, such as "Verified by Visa" two-factor authentication?" - Yes - No Section Practical Strategies and Recommendations 11- What types of additional measures would you recommend to improve security against refund fraud in e-commerce companies? Free Response 12- Employee training and awareness on new online fraud techniques are key aspects of the fraud prevention strategy ? [...]
[...] The suggested solutions proposed a balanced vision, combining both technological tools and operational strategies to strengthen fraud prevention. This part constitutes a crucial moment of reflection, linking theoretical aspects of fraud prevention to concrete and achievable solutions, thus establishing the starting point for the overall conclusions and recommendations that will be discussed in the final part of the study. CONCLUSION The main research objective of this study was to evaluate how artificial intelligence and new technologies can improve the effectiveness of combating different forms of online fraud in e-commerce in France. [...]
[...] The online fraud issue is inscribed in a rich and constantly evolving theoretical framework. Historically, risk management and fraud prevention have been approached from various angles, ranging from internal audit to the development of digital security systems (Chen et al., 2011)5. The first approaches focused mainly on rule-based detection methods, such as identifying unusual transactions in relation to consumption habits (Panigrahi et al., 2009)6. However, these traditional methods have proven to be insufficient in the face of the rapid evolution of fraud techniques, which have become increasingly sophisticated and difficult to detect (Abbasi, Albrecht, Vance & Hansen, 2012)7. [...]
[...] It is this trend that new e-commerce anti-fraud technologies are following. It has become increasingly important to evaluate whether new fraud detection methods integrate into the practical work mode of fraud services or address relevant practical challenges. Multiple Instance Learning (MIL) allows machine learning models to classify entire groups of instances at once, rather than classifying them one group at a time. In many cases of fraudulent use, the fraudulent user targets multiple commercial entities through repeated actions on the e-commerce platform. [...]
