Context-Based Password Cracking for Digital Investigation

Authors: Kanta, Aikaterini

Publication Date: June 2023

Publication Name: PhD Thesis, School of Computer Science, University College Dublin,

Abstract:

Passwords have been the prevailing method of authentication since their inception more than 50 years ago, a trend which has no signs of slowing down in the foreseeable future. Despite alternative authentication methods being developed later, it is reasonable to assume that this prevailing authentication method will not fall out of popularity anytime soon. Passwords are an integral part of the security of digital persons, systems and critical data, and yet, they often remain the weakest entry point to a digital system. The conundrum has driven both the efforts of system administrators to nudge users to choose stronger, safer passwords and elevated the sophistication of the password cracking methods chosen by their adversaries. The system administrator often overcomes the imperfection by skilfully enforcing strong password policies and dutiful password management on the side of the server. But at the end, the user behind the password is still responsible for the password’s strength. A poor choice can have dramatic consequences for the user or even for the service behind, especially considering critical infrastructure. A password itself is indeed an extension of its creator and therefore can be exploited by malicious actors leveraging available contextual information about a target password creator. On the other hand, law enforcement can benefit from a suspect’s weak decisions to recover digital content stored in an encrypted format. Generic password cracking procedures can support law enforcement in this matter – however, these approaches quickly demonstrate their limitations. Recent research has hinted at the influence that context can have on a user during his/her password selection. This information could be of significant added value when digital investigators need to target a specific user or group of users during a criminal investigation. The connection between the password and its creator has given rise to advanced techniques aimed at exploiting user habits for password cracking. Such techniques are often generic approaches that leverage large datasets of human-created passwords. This thesis aims to investigate the hypothesis that bespoke password candidate lists, generated based on available contextual information, can positively impact the password cracking process. For this, a methodology and framework for creating and assessing custom dictionary wordlists for dictionary-based password cracking attacks are introduced, with a specific focus on leveraging contextual information. Furthermore, a detailed explanation of the framework’s implementation is provided, and the benefits of the approach are demonstrated with the use of test cases. This work also introduces techniques for optimising the generation of the bespoke dictionaries, ranking the password candidates in order to maximise the chance of early success. The aim of the proposed approach is to support digital forensic investigators in their criminal investigation – especially when time is of the essence. This approach achieved very promising improvements over existing, traditional approaches in isolation – more than 50 per cent improvement in some instances. This result proves that more targeted approaches can be used in combination with the traditional strategies to increase the likelihood of success when contextual information is available and can be exploited.

Download:

Download Paper as PDF

BibTeX Entry:

@phdthesis{kanta2023PhD,
title="{Context-Based Password Cracking for Digital Investigation}",
author={Kanta, Aikaterini},
school={School of Computer Science, University College Dublin},
month=06,
year=2023,
address={Dublin, Ireland},
abstract={Passwords have been the prevailing method of authentication since their inception more than 50 years ago, a trend which has no signs of slowing down in the foreseeable future. Despite alternative authentication methods being developed later, it is reasonable to assume that this prevailing authentication method will not fall out of popularity anytime soon. Passwords are an integral part of the security of digital persons, systems and critical data, and yet, they often remain the weakest entry point to a digital system. The conundrum has driven both the efforts of system administrators to nudge users to choose stronger, safer passwords and elevated the sophistication of the password cracking methods chosen by their adversaries. The system administrator often overcomes the imperfection by skilfully enforcing strong password policies and dutiful password management on the side of the server. But at the end, the user behind the password is still responsible for the password’s strength. A poor choice can have dramatic consequences for the user or even for the service behind, especially considering critical infrastructure. A password itself is indeed an extension of its creator and therefore can be exploited by malicious actors leveraging available contextual information about a target password creator. On the other hand, law enforcement can benefit from a suspect’s weak decisions to recover digital content stored in an encrypted format. Generic password cracking procedures can support law enforcement in this matter – however, these approaches quickly demonstrate their limitations. Recent research has hinted at the influence that context can have on a user during his/her password selection. This information could be of significant added value when digital investigators need to target a specific user or group of users during a criminal investigation. The connection between the password and its creator has given rise to advanced techniques aimed at exploiting user habits for password cracking. Such techniques are often generic approaches that leverage large datasets of human-created passwords. This thesis aims to investigate the hypothesis that bespoke password candidate lists, generated based on available contextual information, can positively impact the password cracking process. For this, a methodology and framework for creating and assessing custom dictionary wordlists for dictionary-based password cracking attacks are introduced, with a specific focus on leveraging contextual information. Furthermore, a detailed explanation of the framework’s implementation is provided, and the benefits of the approach are demonstrated with the use of test cases. This work also introduces techniques for optimising the generation of the bespoke dictionaries, ranking the password candidates in order to maximise the chance of early success. The aim of the proposed approach is to support digital forensic investigators in their criminal investigation – especially when time is of the essence. This approach achieved very promising improvements over existing, traditional approaches in isolation – more than 50 per cent improvement in some instances. This result proves that more targeted approaches can be used in combination with the traditional strategies to increase the likelihood of success when contextual information is available and can be exploited.}
}