Foster, J. orcid.org/0000-0002-9439-0884, Buckley, A.R. and Zoukas, G. (2026) Adaptive assurance for trustworthy machine learning in the UK energy sector: implications for energy policy and governance. Energy Research and Social Science, 141. 105007. ISSN: 2214-6296
Abstract
The UK's transition to a low-carbon energy system is critically dependent on supporting policies, institutions, and governance. Driven by technologies like smart meters, advanced metering infrastructure (AMI), sensors, and data analytics the digitalisation of energy grids is enabling the development of machine learning (ML) models for energy-related applications. While the integration of ML models into energy grids promises benefits such as more accurate forecasting of energy needs, optimised grid balancing, improved network maintenance, and increased energy efficiency, their use also presents technological, ethical, and governance risks. As a result, the mitigation of these risks through assurance has assumed paramount importance. However, current policy concepts of AI assurance remain generic—universal in purpose and assumed to apply uniformly across all sectors and contexts. This study engages in a theoretical elaboration of the generic concept of AI assurance by specifying its properties, dimensions, and range of variation under UK energy sector conditions. Drawing on Corbin and Strauss's variant of grounded theory and informed by qualitative interviews with ML developers and prospective end-users, it develops a substantive theory of adaptive ML assurance—defined as an assurance process responsive to ongoing changes in context. The study demonstrates that the assurance process shifted from a hierarchical strategy in the early stages of the ML lifecycle to an emergent strategy in the later stages, driven by institutionalised data asymmetries in the UK energy sector. Therefore, the UK-specific institutional conditions at macro and meso levels were constitutive of the assurance process at the micro level, with consequences for the trustworthiness outcome. The study locates the generic concept of AI assurance within its institutional context, revealing what persists and what varies under UK energy sector conditions. The study's findings have several implications for energy policy and governance. To transform energy data from a resource into a strategic asset, energy policy must become more responsive to demand-driven energy grids. This requires a co-constitutive relationship between energy policy and data-driven technological practices, with interventions that increase data availability, improve information governance and interoperability, create incentives for data sharing, and support collaboration and innovation.
Metadata
| Item Type: | Article |
|---|---|
| Authors/Creators: |
|
| Copyright, Publisher and Additional Information: | © 2026 The Authors. Except as otherwise noted, this author-accepted version of a journal article published in Energy Research and Social Science is made available via the University of Sheffield Research Publications and Copyright Policy under the terms of the Creative Commons Attribution 4.0 International License (CC-BY 4.0), which permits unrestricted use, distribution and reproduction in any medium, provided the original work is properly cited. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/ |
| Keywords: | Energy transition; Technological change; Machine learning; AI assurance; Trustworthiness; Energy data sharing; Information governance |
| Dates: |
|
| Institution: | The University of Sheffield |
| Academic Units: | The University of Sheffield > Faculty of Social Sciences (Sheffield) > School of Information, Journalism and Communication |
| Funding Information: | Funder Grant number INNOVATE UK 10094695 TS/Y019792/1 |
| Date Deposited: | 06 Oct 2026 14:01 |
| Last Modified: | 02 Nov 2026 15:08 |
| Status: | Published |
| Publisher: | Elsevier |
| Refereed: | Yes |
| Identification Number: | 10.1016/j.erss.2026.105007 |
| Open Archives Initiative ID (OAI ID): | oai:eprints.whiterose.ac.uk:246076 |
Download
Filename: erss_adaptive assurance_WRR.pdf
Licence: CC-BY 4.0

CORE (COnnecting REpositories)
CORE (COnnecting REpositories)