AI has potential for Britain’s energy sector but more development support needed – report

AI and machine learning have a range of applications within the energy sector but more support is needed for its development, a UK parliament research brief has indicated.

Based on a literature review and input from a range of stakeholders and reviewers, the brief ‘Energy security and AI‘, points to the potential of AI and machine learning to optimise and accelerate energy planning, generation and use.

AI could use data from devices such as smart meters and substation monitoring to help address current regional renewable connection delays and excessive network congestion.

It could also speed up the decarbonisation of the energy system as the UK strives to meet 2030 grid decarbonisation and 2050 net zero targets and reduce costs for consumers.

Have you read?
Tech Talk | ‘No AI without energy’
How genAI can help accelerate the energy transition

However, the brief adds that there are technical and infrastructural barriers to wider adoption of AI in the energy system, including data access, regulation, skills gaps and the availability and reliability of the physical infrastructure that supports AI.

Moreover, stakeholders have raised concerns around data privacy and ownership, cyber security, energy use, fairness, ethical use and operational challenges of using AI in a critical national infrastructure.

Potential mitigation approaches highlighted are the use of privacy preserving technologies, enhanced cyber security and digital literacy, balanced model training and validation and standardised processes and ethical oversight.

But ultimately more support is needed to develop AI in the sector, and regulation needs to change to ensure optimal benefits can be gained from wider integration of AI in the energy system, while avoiding potential risks, the brief states.

Specific policy considerations recommended in the brief are:

• Reframe energy as a shared resource and understand consumer perspectives;
• Update cybersecurity and factor in real-world disaster scenarios, e.g. flooding, as well as supply chain vulnerabilities;
• Consider environmental and infrastructure issues, e.g. energy consumption and scale of data centres;
• Update governance, long-term energy system planning and incentives for data access;
• Build trust, transparency and ethical standards;
• Invest in digital literacy skills and capacity as well as local supply chains;
• Recognise the role of AI as a supporting tool, with a focus on specific applications, such as forecasting and system optimisation.