The TAS Trustworthy and Useful Tools for Mobile Phone Extraction project explores the role AI-driven tools can play in the use of mobile phone data in the criminal justice system. This is an area in which AI can bring significant benefits in terms of efficiency and accuracy. However, it is also a high stakes area in which failures can gravely affect trust. Therefore, it is paramount that we take a responsible approach to the work we do.
As a project team we are very committed to the values of responsible research and innovation (RRI). These values state that care should be taken to ensure the outcomes of innovation align with societal needs and perspectives. This requires a ‘people first’ approach. So rather than asking ‘What new innovations can be developed through cutting edge science?’ We ask: ‘What kind of world do we want to live in, and how can scientific development help us to achieve it?’ Furthermore, in order to do this, it is necessary for us to engage with a wide range of stakeholders so that various responses to this question are represented.
RRI considerations and practices underpin all our project activities. We draw on available resources such as the AREA framework and the prompts and practice cards developed within TAS to help shape what we are doing. For us our RRI approach began during our planning for the project. We co created our research design with our project partners, undertaking anticipation and reflection activities. These identified important dimensions of trust and usefulness that need to be addressed in the exploratory platform for mobile phone extraction we are developing in the project. Consequently, we agreed that our platform, RIME, will be centred around principles of lawfulness, respect for privacy, software quality and usability.
Now that the project is running, we schedule regular RRI sessions for our project team. During these sessions we take time to reflect on our progress and practices, and challenge any assumptions we may have been making. For instance, the findings from our various research activities have revealed organisational tensions across the criminal justice system that can lead to phone owners being pressured or required to hand over more mobile phone data than is relevant to a case (thus creating an intrusion of privacy) as well as tensions that can cause examinations of mobile phone data to be delayed and inefficient. We need to make sure that we accurately represent these tensions and how they affect different stakeholders across the criminal justice system. It is particularly important that we capture the perspectives of those most vulnerable to harm in the current system – such as victims of crime who may be asked to give up their phones and be left without access to them for extended periods of time. Our research outputs will focus on developing digital tools with the capacity to resolve some of these tensions whilst also acknowledging that technical interventions alone cannot fully address the kinds of deeply embedded organisational constraints we have been identifying
Our most recent RRI session has highlighted an issue that we intend to reflect on further in the next few months. Our tool, RIME, is open-source. In addition to being open to those in the community who may seek to develop and modify it, this means it is available to those who wish to inspect it. So, being open-source is an accountability and trustworthiness mechanism. But does it also create some risk? Might our open-source tool make it easier for certain bad actors to download and search through mobile phone data for malicious purposes? As a next step for our RRI activities in the project, we intend to explore this concern further, seeking the informed opinion of experts and identifying whether mitigation strategies can prevent or mitigate this kind of misuse.
You can read more about the TAS Tools for Mobile Phone Extraction project in this blog.
The image above shows one of the Anticipate cards from the RRI Prompts and Practice Cards developed within TAS. It is available here.