Deep opacity undermines data protection and explainable artificial intelligence
Müller VC (2021)
Publication Type: Conference contribution
Publication year: 2021
Publisher: The Society for the Study of Artificial Intelligence and Simulation of Behaviour
Conference Proceedings Title: AISB Convention 2021: Communication and Conversations
Event location: Online
ISBN: 9781713829423
Abstract
It is known that big data analytics and AI pose another threat to privacy, and it is known that there is some kind of ‘black box problem’ in AI. I propose that (1) the ‘black box’ becomes the ‘black box problem’ in a context of justification for judgments and actions, crucially in the context of privacy. (2) This will suggest distinguishing two kinds of classic opacity and introducing a third: The subjects may not know what the system does (‘shallow opacity’), the analysts may not know what the system does (‘standard black box opacity’), or even the analysts cannot possibly know what the system might do (‘deep opacity’). (3) If the agents, data subjects as well as analytics experts, operate under opacity, then they cannot provide some of the justifications for judgments that are necessary to protect privacy – e.g. they cannot give “informed consent” or assert “anonymity”. It follows from (2) and (3) that agents in big data analytics and AI, often cannot make the judgments needed to protect privacy. So big data analytics makes the privacy problems worse, and the remedies less effective. Closing, I provide a brief outlook on technical ways to handle this situation.
Authors with CRIS profile
Involved external institutions
Netherlands (NL)How to cite
APA:
Müller, V.C. (2021). Deep opacity undermines data protection and explainable artificial intelligence. In AISB Convention 2021: Communication and Conversations. Online: The Society for the Study of Artificial Intelligence and Simulation of Behaviour.
MLA:
Müller, Vincent C.. "Deep opacity undermines data protection and explainable artificial intelligence." Proceedings of the AISB Convention 2021: Communication and Conversations, Online The Society for the Study of Artificial Intelligence and Simulation of Behaviour, 2021.
BibTeX: Download