Juniper Research has released a whitepaper, entitled, Beyond Chargebacks: The True Cost of Fraud for Digital Commerce, which examines the current state of the eCommerce fraud prevention market; considering “the impact of evolving digital fraud strategies, including trends like identity theft, account takeovers, chargebacks, policy abuse and friendly fraud.”
The update from Juniper Research reportedly examines various solutions from traditional transaction rules to “emerging technologies such as AI and machine learning.”
Additionally, it includes a forecast of the total “value of fraudulent digital eCommerce goods in 2030.”
As explained in the research report from Juniper Research, chargebacks are a type of transaction reversal that “occurs when a customer disputes a charge directly with their bank or credit card company.”
Chargebacks commonly occur for reasons such “as dissatisfaction with a product or service, fraud, or errors made by the business.”
When a chargeback does occur, the customer’s bank or credit card company refunds the “disputed amount to the customer and deducts the amount from the respective business’s account.”
For many businesses, the chargeback can “result in financial losses; leading to reputational damage, higher fees from payment processors, and even loss of the ability to accept credit card payments.”
The report from Juniper Research further noted that Buy Now, Pay Later (BNPL) is a short-term instalments loan “that lets people pay for their purchases over time.”
Typically, this price will be spread over three instalments, with no interest.
However, some BNPL arrangements allow the customer “to spread the cost of an item from as little as a few weeks to over a year.”
Some BNPL arrangements will “not include any interest for the first 12 months, but after the initial 12-month period, the customer may be charged additional fees and interest.”
BNPL Fraud Methods:
The growth of BNPL services has “expanded the opportunities for criminal exploitation.”
As stated in the research report, fraudsters now commonly will try to commandeer “existing BNPL accounts using synthetic stolen identities.”
As verification processes have deliberately “low friction to protect the customer experience, BNPL has become a particularly attractive target for fraudsters.”
The most prevalent BNPL fraud approaches in 2025 are:
- BNPL ATO: account takeover remains the leading BNPL fraud type. Large volumes of compromised credentials circulate on dark web marketplaces; supplying criminals with easy access to genuine customer accounts. Once inside, they can rapidly make purchases before the legitimate user may realise anything unusual is occurring with their account.
- BNPL Triangulation Fraud: in this scheme, fraudsters advertise goods online at heavily discounted prices. When a customer pays the fraudulent seller, the fraudster will then purchase the actual product itself using BNPL under a stolen or synthetic identity and have it shipped directly to the genuine customer. After delivery is complete, the fraudster raises a dispute with the BNPL provider; taking advantage of poor coordination between merchant and provider systems. This often results in a refund; allowing the fraudster to keep the customer’s payment without ever paying for the goods.
BNPL’s low-friction design usually involves “minimal identity checks and limited use of multi-factor authentication, which makes these attacks easier to execute.”
Criminals frequently obtain login details “through dark web data dumps, phishing campaigns, or credential stuffing attacks, and use them to carry out BNPL fraud at scale.”
The research has revealed the total transaction “value of fraudulent digital goods is outpacing physical goods fraud; rising 162% from a base of $10.4 billion in 2025 to $27.1 billion in 2030.”
The report highlights synthetic “identity use, promo-abuse, and friendly fraud as key drivers behind the surge in digital goods fraud, with fraudsters using AI tools to reach scale.”
Mobile-first purchasing, gaming, streaming, and apps are “widening the attack surface for fraudsters.”
Instant delivery provides a near-zero intervention time; “meaning traditional fraud tools struggle to detect and block fraud before fulfilment.”
The rise of synthetic identity fraud and “credential-stuffing attacks are also enabling sophisticated, high-impact risks.”
Juniper Research’s analysis revealed that the “fastest growing attacks stem from behaviour that appears legit: authorised accounts, valid payment credentials and clean device histories.”
This pattern, fueled by leaking of credentials and AI-driven spoofing, “creates fraud that mimics genuine customers; resulting in systems treating it as benign until fulfilment is complete.”
As such, the report identified the next generation “of fraud defence as shifting from identifying ’bad transactions’ to modelling intent, behavioural deviation, contextual identity, and detecting cross-merchant reputation signals by incorporating advanced AI models.”
The reduced time to intervene “means eCommerce fraud prevention providers must shift to AI-powered, real-time prevention.”
Juniper Research concluded that should these providers fail to implement proactive techniques such as behavioral biometrics, merchants “will drown in increased fraud risks; eroding profitability.”