Skip to main content

Research Repository

Advanced Search

The stuff we swim in: Regulation alone will not lead to justifiable trust in AI

Powers, Simon T.; Linnyk, Olena; Guckert, Michael; Hannig, Jennifer; Pitt, Jeremy; Urquhart, Neil; Ekart, Aniko; Gumpfer, Nils; Han, Anh; Lewis, Peter R; Marsh, Stephen; Weber, Tim

Authors

Olena Linnyk

Michael Guckert

Jennifer Hannig

Jeremy Pitt

Aniko Ekart

Nils Gumpfer

Anh Han

Peter R Lewis

Stephen Marsh

Tim Weber



Abstract

Information technology is used ubiquitously and has become an integral part of everyday life. With the ever increasing pervasiveness and persuasiveness of Artificial Intelligence (AI), the function of socio-technical systems changes and must be considered as playing a more active role. Technology, e.g., in the form of large language models accessed through a chat interface, is now perceived as a social actor rather than as a passive instrument. Therefore, the question of how and when trust in technology and its organisational controllers is well placed is gaining relevance. In this article, we argue that simplistic views of trust that do not reflect the active nature of AI systems have to be replaced with more elaborate models. Regulation alone does not cover the complex relation between human user, AI system, creator, and auditor. We argue that a radical paradigm shift is urgently needed. The current debate that focuses the question of trust on explainable and ethical AI is dangerously misguided. Technology provides the opportunity for some organisations to leverage established prosocial trust relationships and repurpose them for their own narrow interests. The new model suggests an interpretation of socio-technical systems inspired by many-body physics, structuring interactions in a socio-technical system into fields and agents. This naturally explains the perceived agency of AI systems, and leads to actionable recommendations on how the discourse about trust can be reframed.

Journal Article Type Article
Acceptance Date Dec 1, 2023
Publication Date 2023-12
Deposit Date Dec 11, 2023
Publicly Available Date Dec 31, 2023
Journal IEEE Technology and Society Magazine
Print ISSN 0278-0097
Publisher Institute of Electrical and Electronics Engineers
Peer Reviewed Peer Reviewed
Volume 42
Issue 4
Pages 95-106
DOI https://doi.org/10.1109/MTS.2023.3341463
Keywords Artificial Intelligence; Trust; Explainable AI; Regulation; Sociophysics
Public URL http://researchrepository.napier.ac.uk/Output/3420400
Publisher URL https://technologyandsociety.org/technology-and-society-magazine/

Files

The Stuff We Swim In: Regulation Alone Will Not Lead To Justifiable Trust In AI (1.4 Mb)
PDF







You might also like



Downloadable Citations