Ofgem has recently reported on the findings of a public consultation on the scope for ethical artificial intelligence (AI) deployment in the energy sector. This is in response to the Government White Paper in 2023.

The report sets out the key findings arising out of the public consultation, including the key benefits, opportunities and risks from AI adoption in the sector. Here, we summarise the key points.

Core methodologies and outcomes

Ofgem conducted a series of case study orientated consultations with 36 consumer stakeholders across a broad sample range and demographic, aimed to gather diverse perspectives on the use of AI in the energy sector. This included across age ranges, race, nationality, disabilities and overall digital confidence. The report presented several scenarios depicting different potential futures for AI in the energy sector. These scenarios were used to explore the implications of various regulatory approaches and technological developments.

The core findings amongst consumers were, in summary, that AI integration and usage is seen as inevitable but one to be taken cautiously and with oversight:

  • Awareness vs Understanding - Whilst awareness of AI is high, understanding of it is low. Participants in the consultation perceived AI as a ‘hot topic’, frequently discussed in the media and the news. Although most feel they know something about it, a more comprehensive understanding of how it could be used in practice was limited. As a result, immediate associations with AI tend toward extreme examples of how it could be used in the future e.g. a growing use of robots, rather than assimilation of AI in current processes. Negative perceptions and the strength of concerns around AI are particularly heightened for older and low digital confidence participants. For these groups, lower awareness means there is less consideration of the benefits AI can offer and a fear of the unknown.
  • Public Perception – The consultation outlined the prevalent view that broader use of AI is to become more commonplace, with mixed enthusiasm about the impact of AI adoption.  Perceptions are heavily informed by a mix of external factors.  These factors include media stories highlighting the downsides of AI. Negative interactions with tools such as chatbots, where results are inconclusive or feel invasive, also shape views. Participants also had concerns about the threat of job losses stemming from the wider adoption of AI as well as data security risks.
  • AI Opportunities – Participants generally recognised that opportunities exist for the use of AI across all sectors. Common views shared were the role of AI in improving efficiencies and processes for both customers and organisations. These benefits lead to a belief that AI use will be inevitable in key industries.  

AI's Role in the Energy Sector

The Ofgem consultation presented several hypothetical case studies and scenario examples as to how AI may in future be adopted by the energy sector, inviting consumer views as to the opportunities and concerns presented which may reflect wider public opinion:

  • Chatbots/AI Phone Systems – Enhanced AI driven customer service models, primarily to handle basic energy enquiries and initial complaint handling. Automation of these more administrative processes were considered a positive solution for quick and efficient responses in cases of low-risk interactions. The consultation expressed the core concern as to the quality and transparency of the overall service delivery being provided, primarily on the ability of AI to handle complex issues. In these instances, the prevalent view remained for human oversight to manage complex cases and for appropriate accountability.
  • Support for Vulnerable Customers – the consultation also outlined potential AI-powered solutions for assisting vulnerable customers. Ofgem identified a case study for which an AI model can detect signs of financial distress or energy inefficiency in households and proactively help or escalate to human case operators to provide further support.  Concerns included the risk of algorithmic bias associated with consumer profiling.
  • Contestability and Redress Mechanisms – adoption of an AI-powered redress and complaints model allowing for the automation of the initial stages of complaint handling. The aim would be to use AI to help address issues promptly and efficiently. The case study also highlights the need for transparency in AI decision-making processes to maintain consumer trust and produce fair outcomes.
  • Smart Meters – an energy optimisation solution where smart meters equipped with AI capabilities can analyse energy usage patterns and provide eco-friendly and cost-saving recommendations. Such a model may also be integrated with virtual AI-driven energy audits. Whilst the report indicated a largely positive consensus on smart tariffs and smart meters, there were underlying concerns of distrust in AI models which may be subject to error as well as broader transparency concerns on the operation of such a smart meter in practice.
  • Automated Consumer Content – automated consumer emails and updates, personalised for consumers such as energy reports and tips tailored to individual consumption patterns and preferences. Consumer attitudes were identified as being supportive with an assumption that AI advertisement was already the status quo, and that such use cases were largely low risk. Participants expressed some concerns as to the risk of errors in such content or the risk of data privacy breaches.
  • Smart Grid Transformation – optimisation of energy operator grid operations, to enhance the reliability, efficiency and sustainability of the energy grid. Participants recognised the scope for AI use cases in improving wider energy efficiency, but the report noted a general lack of understanding as to the complexity associated with the case study. The report noted a general acceptance and trust in the energy sector to adopt algorithmic solutions for smart grids. 
  • Dynamic Pricing – the recognition of the adoption of AI to provide real time energy pricing models reflecting energy supply and demand. Participants were largely supportive of the adoption of AI for such purposes to reduce energy wastage and to drive sustainability but there were concerns of the risk of an AI model giving rise to unforeseen price hikes. 

If you would like to discuss how current or future regulations impact what you do with AI, please contact Tom Whittaker, Brian Wong, Lucy Pegler, Martin Cook, Liz Smith or any other member in our Technology team.

This article was written by Zac Bourne.