New Study Shows Rapid Increase in Fraud in APAC Region, Highlighting Need for Solutions
North America and Europe have long been hot targets for criminals looking to engage in eCommerce fraud. But, the problem is growing at an alarming rate in other regions, too.
To illustrate, look at the recently released 2024 Identity Fraud Report from user identity verification firm Sumsub. This year’s trend, to borrow the company’s parlance, is the “democratization of fraud.” Thanks to the growth of fraud circles and the proliferation of illicit fraud-as-a-service providers, fraudulent activities are becoming more common, cheaper, and easier to carry out.
Recommended reading
- Address Fraud: How Criminals Swap Addresses to Abuse Victims
- The Top 10 Prepaid Card Scams to Watch Out For in 2025
- How do Banks Conduct Credit Card Fraud Investigations?
- What is Synthetic Identity Theft? How Can Merchants Stop it?
- What is SIM Swapping Fraud & How Does It Work?
- Key Credit Card Fraud Statistics to Know for 2025
This problem is plaguing the Asia-Pacific (APAC) region in particular. On the whole, the region experienced a 121% increase in identity fraud attacks between 2023 to 2024. Some countries were hit much harder than others, though. Identity fraud in Singapore, for instance, surged 207% compared to last year, while Thailand and Indonesia recorded a 206% and 201% year-over-year increase, respectively.
The company identified five prevalent forms of identity fraud observed in the APAC region:
| Identity Fraud Type | Share of Identity Fraud Cases | Remarks |
| Fake Documents | 50% | 70% of fraud within this subcategory involved fake IDs |
| Chargebacks | 15% | -- |
| Account Takeovers | 12% | -- |
| Deepfakes | 7% | Deepfake attacks rose 194% YoY in the APAC region. South Korea recorded a drastic 73% YoY explosion in deepfake fraud. |
| Fraud Networks | 4% | Fraud networks are eight times more prevalent in APAC than in the Americas |
Unfortunately, many developing countries in the region are woefully underprepared to fight the surge in fraud. At 6.02%, Indonesia’s fraud rate is the highest in the APAC region. But, Pakistan and Bangladesh are not far behind, with fraud rates set at 4.28% and 4%, respectively. By contrast, just 1.66% of transactions in the United States are deemed fraudulent.
The takeaway is clear: merchants that do business in the APAC region will experience a rise in identity fraud attacks. Specifically, certain cross-sections of merchants — namely eCommerce merchants operating in the region — could be particularly vulnerable.
What This Means for APAC Merchants
Globally, eCommerce merchants experienced a 137% year-over-year increase in identity fraud attacks, implying an increase in all of the fraud tactics listed above. But, let’s focus on one tactic — chargeback fraud — in particular.
Additional data from Sumsub corroborates the claim that chargeback fraud is on the rise. Per the study, 56% of merchants worldwide experienced chargeback fraud in 2024, a phenomenon that is especially pronounced in the APAC region.
...3.3%, or US$33 out of every US$1,000, of [Asia Pacific merchants’] total eCommerce revenue is lost annually to payment fraud, a significant financial drain. Furthermore, 3.6%, or US$36 out of every US$1,000, of Asia Pacific merchants’ accepted eCommerce orders turn out to be fraudulent, and an additional 5.5%, or US$55 out of every US$1,000, of Asia Pacific merchants’ orders are rejected due to fraud suspicions.
- Visa Small Business Big Future Whitepaper
To make matters worse, merchants who fight fraudulent chargebacks are unlikely to win. As Visa laments, “...these merchants have a dispute win rate of less than 20%, or US$156 out of every US$1,000, highlighting the difficulties they face in contesting fraudulent transactions successfully.”
These figures track closely with the numbers in the Chargeback911® Chargeback Field Report. Our most recent version of the study shows that, on average, merchants win 45% of the chargebacks they represent… but recover revenue only 18% of the time.
Countering AI With AI
One point of emphasis in Sumsub’s Identity Fraud Report is that fraudsters are increasingly relying on AI as a cheap and scalable way to commit fraud. Deepfake attacks, which grew four-fold globally between 2023 and 2024, are created exclusively with artificial intelligence (AI) technology.
That raises a crucial question, though: could the same technology be used to fight back?
Interestingly, yes. Machine learning models, which are a subset of AI, are remarkably effective at detecting fraud. Broadly speaking, machine learning tools, when deployed at checkout, can help merchants collect and analyze vast volumes of payment data in real time.
Unlike traditional algorithms, a machine learning model improves over time. As new data is fed into the model, it improves in accuracy. Once trained, the model can distinguish between normal and suspicious purchase patterns, which can help merchants flag and block high-risk buyers and transactions.
A machine learning model can be deployed to prevent account takeover fraud. For example, it can be fine-tuned to identify failed login attempts, abnormal profile changes, or card-not present transaction patterns that differ from a user’s normal purchase behavior. When detected, the user can be issued an SMS-based multi-factor authentication challenge. Or, the purchase can simply be declined altogether.
Machine learning can likewise curb the use of fake documents by detecting signs of manipulation like forged signatures, pixelated logos, or fake stamps. Data can even be cross-referenced with sources like social media profiles or public records to identify mismatches in account information.
What about deepfakes, though? Well, machine learning can offer solutions here, too. Ai-enabled tools can analyze visuals, audio, video lighting, or even the shape of hands for inconsistencies. Similarly, advanced machine learning algorithms can be deployed in tandem with geolocation and device fingerprinting technologies to stop fraud networks.
Other Advice for What Merchants Should Do
Identity fraud is a broad and multifaceted threat, and no single solution can curb all fraudulent attacks. Instead, merchants need a multi-layered strategy that addresses various pre- and post-transaction threats from different angles. Namely, merchants can:
Adopting a comprehensive fraud prevention strategy is crucial in today's eCommerce landscape. By employing both technological solutions and proactive measures, businesses can mitigate risks and safeguard their revenues effectively.
Implementing robust fraud detection mechanisms is no longer an option. It's a necessity. The trend toward wider adoption of AI and machine learning technology is not going to reverse itself. So, adaptation is necessary. In the end, businesses that prioritize proactive fraud mitigation will invariably stand a much better chance of sustaining long-term profitability.
