Artificial Intelligence in Combating Synthetic Identity Fraud: A Comparative Case Study of Amazon and Shopify E-Commerce

Synthetic identity fraud is a rapidly growing threat, accounting for over $12?billion in losses (?25% of global identity fraud) in 2024 and is projected to rise further. This study evaluates how AI-driven methods detect and prevent synthetic identity fraud on e-commerce platforms. Using a qualitative multiple-case approach (Amazon, Shopify). This study reviews secondary sources including technical documentation, industry reports, and academic literature. The findings of the study show that advanced AI techniques such as anomaly detection, behavioural analytics, and hybrid supervised models have proven to substantially reduce fraudulent activity in the selected cases. The cases selected in this study report fraud reductions of about 30–40% on orders, along with dramatic drops in false positives. However, there are a number of difficulties still experienced in the use of AI, often in terms of false-positives, data availability to train models, and also, the pace of advancement in fraudulent patterns like synthetic identities generated by generative AI. This study recommends continuous refinement of models, oversight of AI decisions by humans (hybrid reviews), and the cooperation of industry stakeholders, through threat sharing intelligence, to aid in adjusting to changing threats.