Real-time Malicious App Detection with AI trained on a 23M+ Global App Database

Background

In today’s world, the primary device for most people is mobile, and phishing exploits this fact to be the most prominent, direct threat to users. The main method for these attacks is malicious apps disguised as legitimate ones. Users grant these apps control over their devices or leak financial information without realizing, leading to real and serious harm.

Traditional detection methods were based primarily on blacklists, meaning they only identify apps that were already registered as malicious. As a result, they have structural limitations when responding to newly disguised apps or modified malware, as they lack of clear criteria for determining whether an app is legitimate.

Key Features

WhiteList Based Malicious App Detection

FakeFinder builds its own WhiteList database based on the latest versions of tens of millions of legitimate apps collected from major global app markets. It first checks whether the apps installed on a user's device are included in this database to determine legitimacy. This is a fundamentally different approach from traditional blacklist-based detection.

Robust Two-Stage Detection

Apps not registered on the WhiteList undergo secondary analysis on the server for malicious actions. Malicious behavior is determined in real time by analyzing installation paths, access permissions, suspicious API calls, behavior patterns, and more. This dual-structure approach enables high detection accuracy even against sophisticated attacks such as disguised or impersonating apps.

Extremely Accurate Detections

FakeFinder is actively deployed within the Korean financial sector, achieving both the highest level of detection accuracy and the lowest false-positive rate. It is a technology recognized and actively used in the most conservative environments.

Lightweight Client API/SDK

FakeFinder can be easily integrated into mobile apps via API or SDK. Since the detection logic runs on a cloud-based server, it uses minimum local device resources.

Case Study: KB Pay

KB Pay offers next-level financial security with FakeFinder.

KB Pay Secures 5,300 Korean Citizens

KB Kookmin Card integrated FakeFinder to its payments application KB Pay, elevating mobile security to a new level.

When FakeFinder detects a malicious app on a user's device, KB Pay immediately blocks both in-person and remote transactions, and notifies the user. As a result, 5300 users were protected just in 2022 alone.

A significant portion of the detected apps were phishing apps impersonating domestic financial institutions or government agencies—high-danger types that could easily lead to financial loss once installed. KB Pay was able to establish itself as a secure and trustworthy payment service based on fast and accurate detection, and proactive responses.

Process

  1. AI-based Database Collects All Legitimate Apps in Real Time

  2. Compare Known Apps with Installed Apps for Legitimacy

  3. Detect and Block all Unknown, Malicious Apps

  4. The Financial App Experience, Safe from Phishing

Experience Everspin

Real-time protection for the web with ever-changing security code and AI-MTD.

AI-MTD security modules for mobile threats.

Combine big data and credit information to root out identity fraud.

Block malicious SMS before they reach users.

Security Consulting: Precise diagnosis with our elite red-team.