Baldini et al. examine in their paper “Ethical Design in the Internet of Things” (2016) a framework on how to cope with privacy risks regarding personal user data in context of the Internet of Things (IoT) evolution. Due to IoT’s rising development and the herewith growing interconnectivity between users and digital systems, data protection is facing more complex challenges. The authors introduce the ‘Ethical Design’ concept, which is an engineering type for IoT applications. In that concept users shall be in control to decide which data they want to disclose in which context when utilizing IoT products. To make these choices available to users, a feasible software architecture framework is needed. Baldini et al. elaborate a framework called Security Toolkit (SecKit), which empowers users to create custom data disclosure guidelines and furthermore to modify them on demand. In addition, the SecKit takes into consideration the varying technical proficiency of users – in relation to the ‘Digital Divide’ – by automating decisions made in the creation of an initial IoT user profile. By making users in control of their privacy policies an additional value is added in this market, which shall partially ensure that IoT supplies are being developed in such a way fulfilling user needs.
Overall the paper is crafted in an organized and comprehensive style of writing. For instance definitions and limitations are clearly stated, finished argumentations are briefly summarized and transitions in topics being discussed are explained at first.
However looking at the paper in detail, it is apparent that Baldini et al. have a frequent tendency to craft very long sentences. For instance on page 19: “The SecKit policy-based framework presented in this paper can give more control to the user and it can automate some of the complex decision processes in the interaction of the user with the IoT, but a complementary set of regulatory measures and best practices could make the application of SecKit more effective.” Breaking the sentence into two when the counter argument is introduced (“,but”) and adding an “also” to support the relation to the sentence before would have made this phrase more comprehensive: “The SecKit policy-based framework presented (…) in the interaction of the user with the IoT. But a complementary set of regulatory measures and best practices could also make the application of SecKit more effective.”
Another point of critique is the economical perspective the authors incorporate. Quoting from page 5: “To address this issue, and while acknowledging that a legal framework alone will not be able to ensure this balance, we propose in this paper the expression ‘‘Ethical Design’’ for future IoT devices and services, where ‘‘ethical choices’’ made available to users within the digital architecture become an added value that users are willing to pay for, and that will ensure the needed balance.” Baldini et al. don’t have an economical background, but come from the areas computer science, public administration, service design and STS. By using the phrase “(…) added value that users are willing to pay for” they incorporate economical terms. However little evidence or limitations are stated at this passage that this added value will be a sufficient incentive to bring an ethical design concept to market acceptance. Therefore this argument falls short because customer demand for an added value is taken for granted, not taking into account other types of consumer behaviour (for instance negative or unwholesome demand). The authors could have bypassed this by changing the phrase highlighted before into a subjunctive “(…) added value that users could be willing to pay for”.
Lastly, I want to criticize the overall use of abbreviations when being mentioned multiple times. While some abbreviations such as IoT, ICT etc. are commonly known in the scientific arena, others such as PET, ECA, PDP etc. require detail knowledge. Even though mostly all abbreviations have been written in full when being introduced at first, the reader probably won’t be able to remember them all throughout the text. Forcing the reader to go back to the initial introduction of an abbreviation minimizes the text’s reader-friendliness. Of course when phrases are being mentioned very frequently and are an essential part of the argumentation, such as ‘SecKit’, it can also make sense to use commonly rather unknown abbreviations.