Type to search

Mime Stories Startup News Startups

Grab Uses AI to Detect Fraud Behavior and Fake Accounts

Grab Uses AI to Detect Fraud Behavior and Fake Accounts

Tech company, Grab sees safety as important. Apart from the safety of users and driver-partners, Grab also views online safety as a priority. Especially in pandemic conditions, online fraud is on the rise and perpetrators are increasingly targeting both individuals and institutions.

Especially for Grab, which operates to serve people in Southeast Asia, it appears that the level of digital literacy in this region is still low. Therefore, apart from technology, education is one of the things that is being intensified.

Grab Uses AI to Detect Fraud Behavior and Fake Accounts

Head of Technology, Integrity Group, Transport, and Patents Office Grab, Wui Ngiap Foo, admitted that Grab is not free from various online attacks.

“Online attacks range from the use of GPS to fake activities to the theft of OTP (one-time password) with social engineering,” said Wui, in a media session held through the Zoom platform, Wednesday (11/18/2020).

Wui admitted that the methods used for social engineering are getting smarter, the goal is to get OTP and access the user’s account. Wui said, Grab would never ask for OTP or personal data with the lure of gifts or to provide assistance.

Invest in Technology to Protect Users

For this reason, in addition to investing in technology, Grab also seeks to increase user awareness and education. So that users can protect their accounts. Wui said, Grab collaborated with law enforcement in various market countries, including in Indonesia and Singapore because he considered fraud is not only the responsibility of one person.

The company implements AI-based face recognition for driver-partners before receiving orders. From time to time, Grab can also ask drivers to take a selfie for verification. Passengers who are using the Grab service for the first time are also asked to take a selfie.

Grab Uses AI to Detect Fraud Behavior and Fake Accounts

Even so, Foo guarantees that biometric data and others are safe. Also, AI is embedded in features that are active during the trip. The features referred to are transaction status, GPS, traffic conditions, maps, and telematics.

Artificial Intelligence and Machine Learning

Furthermore, to increase security, Grab is also investing in artificial intelligence and machine learning. “Because AI is the only way to keep up with the number of attacks and fraudulent practices. We try to stay one step ahead of fraudsters. AI is our major tool for dealing with security incidents online or offline,” he said.

Wui said that the fraud was complex, starting from people hiding behind user accounts, creating fake accounts with the names of one of the users, to some pretending not to be guilty.

“What we are trying to do is protect our users. We are trying to understand the journey behavior of customers,” he said. He gave an example, if someone is a normal user, Grab can see their browsing history in the application. For example, if you want to order a vehicle, real users will check rates or prices.

Grab Uses AI to Detect Fraud Behavior and Fake Accounts

Or if you want to order food, normal users will go back and forth to check which food you want, whether it’s a burger or Thai food. All of that is recorded in history which can only be accessed by Grab.

Map Real Users and Fraudsters

Wui said that this method is an AI-powered behavior modeling. From this browsing behavior, Grab can map and determine how real users behave and how fake users log into their accounts, then will immediately take the money stored in the wallet.

“From there, it will be a behavioral graph and an assessment can be generated. Based on that, we can take steps, for example limiting cash flow and others,” he said.

Going forward, Grab will also launch QR Code authentication to ensure digital security for transactions on the desktop, where users need to scan the QR code as an additional layer of security. This is intended to ensure the security of transactions made by the users.