Virtualization-based Penetration Testing Study for Detecting Accessibility Abuse Vulnerabilities in Banking Apps in East and Southeast Asia

Wei Minn 1 , Phong Phan 2 , Vikas K. Malviya 3 , Benjamin Adolphi 4 , Yan Naing Tun 1 , Henning Benzon Treichl 5 , Albert Ching 2 , Lwin Khin Shar 1 , David Lo 1 · 2026

Read Virtualization-based Penetration Testing Study for Detecting Accessibility Abuse Vulnerabilities in Banking Apps in East and Southeast Asia by Wei Minn 1 , Phong Phan 2 , Vikas K. Malviya 3 , Benjamin Adolphi 4 , Yan Naing Tun 1 , Henning Benzon Treichl 5 , Albert Ching 2 , Lwin Khin Shar 1 , David Lo 1 (2026) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2601-21258.txt.

Introduction

Wei Minn and eight co-authors examine a weak spot in Android banking apps used across East and Southeast Asia. Mobile banking lets people handle money from their phones, which is fast and handy. The same apps draw criminals who hunt for flaws that expose private financial data.

The study uses virtual setups to test the apps the way an attacker would. It focuses on abuse of features built to help users with limited vision or movement. Those access features can open side doors into an app if they are not guarded well. The work shows banks where such flaws hide and how they can be found.

Catalog record

CREATORWei Minn 1 , Phong Phan 2 , Vikas K. Malviya 3 , Benjamin Adolphi 4 , Yan Naing Tun 1 , Henning Benzon Treichl 5 , Albert Ching 2 , Lwin Khin Shar 1 , David Lo 1
YEAR2026
COLLECTIONPAPERS
FORMATTXT
FILE SIZE26 KB
VERSION1.0
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2601-21258

Browse the full library · About the Tower of Babel

Full text

The opening of the work, from the library's hosted text:

Virtualization-based Penetration Testing Study for Detecting Accessibility Abuse Vulnerabilities in Banking Apps in East and Southeast Asia Wei Minn 1 , Phong Phan 2 , Vikas K. Malviya 3 , Benjamin Adolphi 4 , Yan Naing Tun 1 , Henning Benzon Treichl 5 , Albert Ching 2 , Lwin Khin Shar 1 , David Lo 1 Abstract Android banking applications have revolutionized financial management by allowing users to perform various financial activities through mobile devices. However, this convenience has attracted cybercriminals who exploit security vulnerabilities to access sensitive financial data. FjordPhantom, a malware identified by our industry collaborator, uses virtualization and hooking to bypass the detection of malicious accessibility services, allowing it to conduct keylogging, screen scraping, and unauthorized data access. This malware primarily affects banking and finance apps across East and Southeast Asia region where our industry partner’s clients are primarily based in. It requires users to be deceived into installing a secondary malicious component and activating a malicious accessibility service. In our study, we conducted an empirical study on the susceptibility of banking apps in the region to FjordPhantom, analyzed the effectiveness of protective measures currently implemented in those apps, and discussed ways to detect and prevent such attacks by identifying and mitigating the vulnerabilities exploited by this malware. I Introduction Android applications have reshaped the banking landscape. Users can now conduct diverse financial activities through a simple tap on their smartphones or tablets. Furthermore, these apps also provide advanced functionalities such as budgeting tools, expense tracking, and tailored recommendations, which help users make well-informed financial choices and control their finances. Security issues have also emerged with these substantial changes in the banking sector. The popularity of these apps has also drawn cybercriminals who aim to make money by taking advantage of security flaws in the app. Due to the sensitive nature of financial data, strict security measures are required to guard against fraud and unauthorized access. FjordPhantom [ 6 ] is one such malware that utilizes virtualization to target Android applications. It was identified by Promon [ 16 ] , a Cybersecurity firm and our industry partner. FjordPhantom uses virtualization to disable the detection of malicious accessibility services. This allows it to operate without being flagged by security software. FjordPhantom operates in two main stages. The first stage involves deceiving users into installing an app that seems benign. This app contains the code to disable the detection mechanisms for accessibility services. This is often done through social engineering tactics, such as fake updates or disguised apps. Attackers must also convince the user to activate the malicious accessibility service, which grants the service significant permissions. In the second stage of the attack, the FjordPhantom takes advantage of the disabled detection to perform malicious activities such as keylogging, screen scraping, or other forms of unauthorized access to the app’s data. Motivated by the need to protect the clients of our industry partner from the threat of FjordPhantom in the mobile banking domain, we evaluate banking Android apps and report our findings in this paper. More specifically, we selected banking apps from the East and Southeast Asia where the majority of the clients of our industry partner are based in. Then, we developed a penetration testing system for exploring vulnerabilities of these apps, and analyzed the results of the attacks. To ensure the testing incurs no infringement of privacy or any form of damage to any person or organization, we put strict limits on the scope of the attack to local device only, and involves only a rudimentary disabling of accessibility service detection in the apps. Though limited in scope, the penetration testing adequately demonstrates the apps’ vulnerability to FjordPhantom while avoiding any ethical implications. We also analyzed different protection measures that are used by the banking industry in their apps and checked their effectiveness against FjordPhantom. We then discussed adequate preventive measures to detect and patch the vulnerabilities that allow malware like FjordPhantom to successfully attack the app. Lastly, we discusses how these measures can be incorporated into prevailing software development practices in the industry. The remainder of the paper is structured as follows. Section II describes FjordPhantom – its characteristics and how it works. Section III details the experiment setup to evaluate vulnerability of selected banking apps and protections they deploy. Section IV discusses how banking apps can be vulnerable against FjordPhantom and the current state of the distributions of those vulnerable apps, and discusses the insights drawn from this study and the implications for practitioners. Section V presents how FjordPhantom can be prevented. Section VI provides a summary of the related work. Section VII provides concluding remarks and future work. II FjordPhantom Attack The goal of the creators of FjordPhantom is to perform attacks based on malicious accessibility services (e.g., to obtain user input from the app, perform overlay attacks, or inject touch events) on a series of banking apps. However, since this is a very common attack performed by malware, banking apps typically detect malicious accessibility services and either warn the user not to proceed to even disallow the usage of the application altogether. Because of that, the creators of FjordPhantom needed to find a way to work around these protections. Typically, this is done using a repackaging attack where the logic to detect accessibility services is removed statically and the application is recompiled. This does, however, violate the integrity of the application and is commonly detected by anti-tampering measures already deployed inside banking apps. The authors of FjordPhantom therefore bypass the detection of malicious accessibility services without tampering the app by using a virtualization solution that enables them to run the target app in a virtualized environment that they control. In that environment, they can place runtime hooks into the functionality that is responsible for detecting malicious accessibility services without needing to modify the application, thus circumventing the anti-tampering mechanisms. As shown in Figure 1 , FjordPhantom relies on various components for attacking apps discussed below. Fig. 1 : FjordPhantom Attack App Virtualization. FjordPhantom makes use of the application virtualization solution which is an app (host) that hosts different apps (plugins) to run in the context of the host app. To achieve this, application virtualization solutions encapsulate the system requests of the plug-in applications and act as the proxy between the plug-ins and the Android system [ 4 ] . This breaks the Sandbox security feature that restricts different apps and services from accessing each other’s file system and memory because the plugins’ requests are encapsulated by the host’s process ID. Figure 2 depicts an example permission escalation scenario where a plugin can access sensitive data after inheriting permissions given to the host app by the Android system. Exploiting this concept of encapsulation and sharing the same process ID, any app (like FjordPhantom) that hosts plugins (and uses the virtualization solution) have access to the memory and file systems of the plugin apps. Fig. 2 : Broken Sandbox due to App Virtualization API Hooking. Next, FjordPhantom uses a Java API hooking framework to intercept and modify API communication between the target app and the Android system during runtime. Hooking allows FjordPhantom to avoid having to statically modify the original code which would trigger anti-tampering hardening

Read the complete text