Efforts to establish common approaches to testing and governing artificial intelligence in energy infrastructure are being taken forward by a new Energy Institute (EI) industry committee, amid concerns about the challenges of deploying AI in safety-critical operational environments.
The Applied AI for Energy Committee, which held its inaugural meeting in London on 15 October, has identified two initial priorities: developing an AI governance framework for the energy sector, and establishing common approaches to AI evaluation, assurance and benchmarking.
The initiative brings together energy companies, technology providers and AI specialists to address technical challenges that individual organisations may struggle to resolve independently.
AI is already being used in areas such as predictive maintenance, engineering and operational optimisation. However, extending its use into more critical applications raises questions about how systems should be evaluated, how their performance can be verified and who remains accountable for decisions informed by AI.
Dan Jeavons, President of Applied Computing, has been appointed Chair of the committee, with Andy Webster, Senior Director of Digital and AI Transformation at KBR, serving as Vice Chair.
“AI is already finding its way into control rooms across energy, but the sector is still working out, company by company, how to test it and trust it,” said Jeavons.
“The Applied AI for Energy Committee gives the industry a neutral place to build that evidence together, so the people running safety-critical assets can adopt AI with confidence.”
From experimentation to deployment
The programme builds on an EI-convened industry forum on sovereign AI for UK energy security, held in July. Its findings, published in a new report, identify evaluation and procurement, professional competence, accountability, data rights and operational assurance as priorities for further work.
The report suggests that while AI is already delivering benefits in some applications, moving from individual projects and demonstrations to wider operational deployment remains difficult.
Martin Maeso, EI Director of Technical & Innovation, said the challenges surrounding assurance, accountability, skills and data could not be addressed effectively by companies working independently.
The committee intends to develop practical industry resources, including good practice guidance and training, concentrating on areas where cooperation does not compromise commercial competition.
Its work will form part of the EI’s wider Technical & Innovation Programme, with the Institute providing secretariat support.
The first progress update is expected at the AI Summit during the Energy Institute’s International Energy Week in London, on 23–25 March 2027.
No timetable has yet been announced for publication of the initiative’s frameworks or findings.