Towards automated digital alibis

Abstract

In recent years, society is using media and communication technology more. One consequence of this is that there is more opportunity for its abuse, as well as use in crime. If there is more criminal usage of digital communication technology, we need to reliably be able to use digital evidence in both prosecutions and criminal defence cases. Digital evidence has already been used in a number of case decisions. The main focus of this study is to show that forging of digital alibis is possible, which can impact on the reliance on digital evidence. The creation of a digital alibi using automation is the focus of other studies, and we are also looking at: social interaction during the simulation of a user's behaviour, and whether behaviour can be simulated realistically and with limited computer evidence of the simulation. For this purpose an alibi framework is created that mimicked a typical user's behaviour. This automation is intended to be implemented in such a way that it may not be possible to decide whether the traces of exercising actions recorded on the computer originate from either: a human user or a simulated user. The alibi framework is adaptable to users' behaviour and preferences. Furthermore, forensic analysis is applied to verify whether the framework leaves any traces behind that could indicate the usage of a simulator. Social interaction is simulated by using a chatbot which used email and messaging and analysed the computers for their performance in creating a digital alibi. The performance criteria for the framework was based on metrics taken from real person computer behaviour. By evaluating the forensic analysis of simulated user's hard drive and the real persons' behaviour, it turned out that the simulator fits in the average range of real persons for nearly all categories. Therefore, simulation of user's behaviour is possible, on condition that determining the sources of digital evidence is hard to verify. However, having a closer look at social communication, there is room for improvement, e.g. the chatbots' text comprehension and email content.