How artificial intelligence prevents payment fraud

September 14, 2017 08:00
While online shopping is getting increasingly popular, cyber security has become an important concern, says Erwin Huang (above), associate professor and senior adviser (entrepreneurship) of HKUST. Photo: HKEJ

The e-commerce industry has continued to gain traction in Hong Kong in recent years. About 88 percent of Hong Kong people shopped online last year. Most shoppers were aged below 30, according to a report from Nielsen Corp.

About 41 percent of these online shoppers used mobile devices. Clothing (52 percent) was the top category of online purchases in the city, followed by air tickets (41 percent).

Surprisingly, the average order value of online shopping orders made by Hong Kong people actually doubled the amount of their counterparts in the United States. The amount they spent in clothing was four times that spent by Americans.

While online shopping is getting increasingly popular, cyber security has also become a new concern. Fortunately, tremendous advancement in machine learning and artificial intelligence in recent years have helped identify suspicious transactions and prevent payment frauds, according to a technology expert.

“Online shoppers should always be aware of cyber security threats. People who use public free Wi-Fi to shop online may easily fall victims to the hotspot scam,” said Erwin Huang, associate professor and senior adviser (entrepreneurship) of the Hong Kong University of Science and Technology.

“Hackers may have set up bogus hotspots offering free Wi-Fi in public cafes and stolen mobile users’ passwords, credit card numbers and personal details.”

Huang said users should rely on 4G mobile networks if they are unclear about the Wi-Fi source. He said mobile users can also use a second level authentication.

“Simply speaking, when you type in your login and password, the merchant will send you a string of codes through a text message to complete the transaction,” he said. “Such procedure can ensure that no one can access your account even if hackers have stolen your password.”

However, consumers may find it inconvenient and frustrating as they have to input more information during the checkout process.

Four ways to keep transactions safe

To ensure security for online payments, Visa organized the Visa Security Summit 2017 in South Korea earlier this year, with an aim to announce and underscore a number of initiatives in improving the user experience and payment security.

In general, there are four ways to ensure payment safety:

1. Protect data

2. Devalue data

3. Harness data

4. Empower consumers.

Under the Payment Card Industry Data Security Standard, institutions are required to encrypt cardholders’ information that they have collected, processed, transmitted and stored. Token technology is used to protect credit card information by replacing the card number with unique tokens so that hackers will not be able to use stolen account information to make counterfeit transactions.

Visa transaction controls enable cardholders to control their credit card usage settings. For example, they can suspend overseas spending, stop charging for online games, or set a spending cap on online transactions. These settings empower consumers to mitigate their risks of fraud.

As for the card-not-present transactions, Visa also has introduced 3-D Secure 2.0, which helps issuers and merchants distinguish good transactions from bad, while allowing transactions to continue to happen at lightning speed.

The new 2.0 version of the technology enables a real-time, secure, information-sharing pipeline that merchants can use to send an unprecedented number of transaction attributes that the issuer can use to authenticate customers more accurately without always asking for a static password or slowing down commerce.

A system to learn customer behavior

3-D secure 2.0 is seen as one of the advanced transaction authentication methods adopted in payment security, Huang said.

“For example, if I am in Hong Kong in the morning but an online transaction is detected in Paris in the afternoon and another one in San Francisco in the evening, artificial intelligence can easily identify that these are payment frauds,” Huang said.

Recording customer habits does not only help ensure security but also helps provide superior experiences to consumers, Huang said.

“The emerging trend of experimental consumerism will have a big impact on marketing,” he said. “When you open the Airbnb app, you will see a lot of different experiences rather than hotel options, such as learning Jazz in Cuba, spending a few days with wolves in Anacortes in US. Artificial intelligence will learn your preferences when you interact with the system. If you like to celebrate your birthday in Paris every August and consume some wine in a hotel room with a view of the Seine River, the system will automatically provide you some relevant information prior to your birthday.”

“You won’t find these advertisements annoying, and you won’t even check the prices before you make the purchases,” Huang said, adding that nowadays consumers care more about user experience than sales discount.

It’s reported that 3-D Secure 2.0 will be launched in Hong Kong early 2018.

-- Contact us at [email protected]


Hong Kong Economic Journal