Towards an accountable Internet of Things: A call for reviewability

Abstract

As the IoT becomes increasingly ubiquitous, concerns are being raised about how IoT systems are being built and deployed. Connected devices will generate vast quantities of data, which drive algorithmic systems and result in real-world consequences. Things will go wrong, and when they do, how do we identify what happened, why they happened, and who is responsible? Given the complexity of such systems, where do we even begin?This chapter outlines aspects of accountability as they relate to IoT, in the context of the increasingly interconnected and data-driven nature of such systems. Specifically, we argue the urgent need for mechanisms-legal, technical, and organisationalthat facilitate the review of IoT systems. Such mechanisms work to support accountability, by enabling the relevant stakeholders to better understand, assess, interrogate and challenge the connected environments that increasingly pervade our world.

A Conceptual Evaluation of the Data as Partner (DAP) Framework

“…The emphasis on proactively addressing ethical, legal, privacy, and bias risks highlights that the DAP paradigm cannot overlook principles of responsible data management in the pursuit of close datadecision partnerships. As highlighted by Cobbe [41], failing to assess risks of anthropomorphising data can propagate harms. The framework requires robust ethical guardrails, not just technical guidance.…”

Making Algorithmic Systems More Meaningfully Transparent for Users Through Visualisation

“…Moreover, the relation between the concepts and implementation of transparency, visualisation, and design focussing on users' interests and wellbeing has not been systematically explored yet. Our approach builds upon the extensive literature in HCI on visualisation from the perspective of users [7][8][9][10], as well as technical and interdisciplinary work on the opportunities, challenges, and issues related to transparency and explainability [5,6,13,14,18,25,26]. Our work also aims to contribute to the latest research in design for wellbeing with a specifc focus on visualisation [11,12,21,22].…”

“…To become meaningful, transparency about systems has to be contextually appropriate, that is relevant and proportionate to the users' information needs and levels of understanding, accurate, i.e. correct, complete, and representative, as well as comprehensible for the users [18]. Our research proposes to respond to the following question in HCI: how to make algorithmic system transparency more meaningful to users with visualisation [5]?…”