Over the last decade and a half, online shopping has become a way of life and has transformed the way people around the world purchase things. According to Statista, with over 5.5 billion internet users globally, the number of online shoppers continues to rise steadily. In 2025, global retail e-commerce sales are projected to surpass 4.3 trillion U.S. dollars, with expectations for even greater growth in the years ahead.
Meanwhile, this stupendous growth has also brought a surge in fraudulent activities that threaten the security of online transactions. Apart from causing great losses for online merchants, these activities also dent the confidence of online buyers. A very well-known example of e-commerce fraud was in the year 2013, when cybercriminals gained access to Fortune 500 retailer Target’s network through a third-party Heating, Ventilation, and Air Conditioning (HVAC) vendor, and stole credit and debit card information of about 40 million customers. This incident was a stark reminder that no organisation, whether small or large, is immune from cyber-attacks. Fraudsters are becoming extremely sophisticated and using more and more innovative ways to target online buyers and ecommerce merchants.
However, it is not all bad news. Artificial Intelligence (AI) has appeared like a saviour for ecommerce and is playing a huge role in curbing the menace of online fraud in Ecommerce.
how fraudsters exploit online shoppers and merchants
There are a number of ways in which fraudsters target online shoppers and merchants who are part of the ecommerce ecosystem and steal money, goods and customer information. Fraudsters use stolen credit card information to make purchases without the knowledge or permission of the cardholder. Chargeback fraud is when a customer buys and receives a product but claims non-receipt of the product and demands a refund. Phishing and social engineering is another type of common fraud where shoppers reveal sensitive information to fraudsters without realising their malicious intent.
AI, the bane of ecommerce fraudsters
AI has been a real game-changer in the area of ecommerce by helping catch and prevent fraud. What’s more, AI systems keep up with evolving fraud tactics by learning and improving with each new fraud attempt.
real-time alerts and anomaly-detection – Static patterns and rules to catch online fraud are not very effective. On the other hand, AI has the ability to catch subtle changes in patterns and flag the same for further investigation. It does this by viewing very large data sets and analysing transactions in real-time. It can also compare current data with historical data – which helps in spotting anomalies or outliers which could indicate potential problems in the transactions.
behaviour pattern changes – It is interesting to note that AI is actually able to study the behaviour of customers making purchases. It does this by learning what is normal behaviour for each customer. Thus, any unusual or suspicious activities, such as logging in from a new country or IP address, logging in from a different device, changes in the frequency of logging in or making high-value purchases without the usual browsing time are noted by the system and flagged.
multi-layered authentication – While making purchases online, online stores check the identification of the purchaser through passwords. AI helps by adding an extra layer of protection through multi-layered authentication. This could be through biometric authentication, such as facial recognition and fingerprint scanning, which makes it more difficult for fraudsters to gain unauthorized access. Such multiple layers of authentication ensures that fraudsters need to beat more barriers before they can commit any kind of fraud.
adaptive fraud scoring - AI assigns a risk score to each transaction. If the transaction is low risk, it goes through without any hindrances. However, high risk ones are flagged or blocked for fraud prevention. A case in point is Paypal, one of the world’s largest online payment platforms. The company examines over 500 data points across each transaction and assigns a risk score. This approach has helped the company block $500 million in fraud quarterly, and maintain fraud rates well below the industry average while delivering smooth customer experiences.
chargeback and refund abuse detection
With AI scanning millions of transactions, the models are able to identify patterns of abuse in returns, refunds, and chargebacks. They can recognize when the same customer keeps requesting refunds and chargebacks. Traditional fraud filters often block legitimate customers. AI helps in reducing false positives, minimizing unnecessary friction and thus protecting revenue.
the way forward
“For ecommerce, the most important thing is trust.” – Jack Ma, Chinese entrepreneur and the co-founder of Alibaba Group
Having ecommerce fraud prevention isn’t just about stopping bad actors. It is also about protecting an organization’s brand and keeping the payment infrastructure healthy. Studies have revealed that ecommerce frauds are projected to surge from $44 billion in 2024 to $107 billion in 2029; a staggering growth of 141%! To combat this rising tide of fraud, companies need to take adequate steps to embrace AI to quickly identify threats, reduce fraud losses and maintain valuable trust in the digital market place.
how Infosys BPM can help
With the surge in fraudulent activities, intensified by rapid digital transformation and pandemic-driven shifts, retailers must prioritize robust eCommerce fraud management solutions to safeguard against malicious transactions. Leveraging advanced AI, future-ready technologies, and the expertise of seasoned fraud detection professionals can significantly impact overall profitability. Infosys BPM delivers acomprehensive retail and eCommerce fraud management solution, built on years of cross-industry experience worldwide and tailored to meet each client’s specific needs.
