These data are related to paper"An Empirical Study on Malicious SMS-Based Apps".
In this work, we summarize 7 fine-grain categories for SMS-based malware and the general attack chain for each category. To contribute to relevant research topics, such as AI-based interpretable malicious behavior classification, we release 56 malware samples and their corresponding attack chains as a benchmark. We hope they can be used as a fundamental dataset in future SMS-based malware research.
The following figure shows a brief description about the potential harm,characterization-based family label, discoverability, and the number of collected cases.
The following is a brief introduction about the 7 categories and their corresponding attack chains.
This category of malware can steal victim’s messages by secretly sending them from the infected device to a remote server.
This category of malware can cause unaware service charges by bypassing a two-factor authentication (2FA), such as typing in a verification code by user, and completing a subscription payment.
This category of malware can cause unaware service charges by bypassing a 2FA and completing a payment, too. However, different from “Misclick to pay”, the malicious behavior will be triggered directly after started.
This category of malware can steal victim’s privacy, such as the account information and login password to digital property, by providing a fake login/payment interface.
This category of malware can cause unaware service charges by bypassing 2FA and completing a payment.
This category of malware can perform SMS-based malicious behaviors with remote control commands.
This category of malware can be spread to more devices efficiently by sending the download link to all contacts on victim’s device.
