This week, Mastercard launched Threat Scan, a global solution to help banks spot authorization weaknesses before a breach occurs. This new product is in line with recent cybersecurity trends toward identifying threats before they occur.
In August, Visa launched new early detection and warning tools that use artificial intelligence (AI) and deep learning across the company’s transactional records to identify anomalies that indicate that hackers may have gained access to a merchant or financial institution’s systems and are gearing up to use them for malicious purposes.
The latest Digital Fraud Tracker details established and emerging fraud trends and focuses on how financial institutions and merchants can use machine learning (ML) and AI to become more proactive in detecting and fighting digital fraud.
Here are some recent statistics about different types of digital fraud.
Account Takeover (ATO) Fraud
Account takeover (ATO) fraud shows no signs of letting up. According to the latest DataVisor Fraud Index Report, fraud losses hit $14.7 billion last year, with ATO fraud contributing $4 billion in losses. The eCommerce sector was a favorite target for fraudsters, representing 40 percent of total losses due to ATOs.
Online Shopping Fraud
As alluded to above, eCommerce is a popular fraud target. The Federal Trade Commission’s new Consumer Protection Data Spotlight shows that millennials (ages 20 to 39) are twice as likely to report losing money to online shopping fraud than their older counterparts. Online shopping fraud reports include complaints about items that are never delivered or are not as they were advertised. The top five frauds to which millennials report losing money are online shopping frauds, business imposters, government imposters and fake check scams.
As reported recently, a new strain of ransomware has been hitting U.S. firms and firms based overseas. The latest news centers on Sodinokibi, a ransomware strain that has helped fraudsters make higher ransom demands.
Cybersecurity firm Barracuda found that more than 50 ransomware attacks have been reported in the first half of 2019, and a report from cybersecurity solution provider Coveware discovered that the cost of ransomware attacks is increasing. The average value of ransom payouts during Q2 2019 was $36,295, up significantly from the $5,973 reported in Q3 2018.
Because of the huge number of emails sent daily — by some estimates, 3.7 billion people send around 269 billion emails every single day — this type of online communication is vulnerable to phishing attacks. Recent research indicates 30 percent of targeted attempts are made via phishing emails, and that 15 percent of victims are repeat targets.
Machine Learning and Artificial Intelligence as Proactive Solutions
Banks are increasingly employing AI and ML tools to fight fraud. Recent data found that 28 percent of banks are prioritizing security improvements in 2019, and 77 percent of banks are already putting AI solutions to use.
A recent Visa survey of professionals from the banking, FinTech and merchant sectors found that 62 percent expect new payment methods to become available over the next two years.
However, 61 percent are concerned that these new technologies will make them more vulnerable to fraud. Specifically, 68 percent are concerned about mobile banking fraud and 58 percent are concerned about peer-to-peer (P2P) payments.
More than three-fourths (77 percent) are ready to invest in solutions. The report recommended adoption of advanced solutions like AI and ML to make it easier to detect and analyze fraud.
Legacy fraud solutions are limited because they can only react to incidents after they unfold, an approach that is no longer effective given fraudsters’ creativity and adaptability.
Andrew Sloper, head of digital authentication at JPMorgan Chase, spoke to PYMNTS about how ML is enabling the bank to take a preventative approach to fighting fraud by analyzing data in real time and finding activities that could point to more serious fraud threats. This reverses the traditional model of responding to fraud after it is detected. “It essentially allows us to detect patterns as they emerge, even ahead of fraud being committed,” he said. “It helps [us] to take a more proactive approach,” he said.
Yinglian Xie, co-founder and CEO at DataVisor, also believes in the power of AI and unsupervised machine learning to become more proactive rather than reactive to fight fraud. “Advanced unsupervised ML solutions can find correlated patterns that reveal coordinated malicious activity, enabling businesses to stop new and emerging attacks before they launch,” she said.
Unsupervised ML solutions promise many benefits, but financial institutions and merchants looking to implement them may perceive them to be costly investments. However, according to PYMNTS’ AI Innovation Playbook: Perception Versus Reality in Payments and Banking Services, 61 percent of FIs said they were planning to invest further in their supervised and unsupervised learning tools.