Unlocking digital innovation in the electricity system

1. Introduction

Electricity systems are an increasingly important part of the wider energy landscape, with almost half of all energy investment now directed towards electricity and end-user electrification (IEA, 2025[2]). Moreover, electricity systems are now a lynchpin for the continued progress of overall digital transformation across countries given their fundamental role in powering digital technologies, from the fabrication of key semiconductors required for artificial intelligence (AI) systems to develop, to the data centres where AI systems are trained and where inference takes place (IEA, 2025[2]; OECD, 2023[3]; OECD, 2022[4]). However, like other industries and sectors, electricity systems themselves are being transformed by the innovative use of digital technologies, notably AI. The efficiencies and improvements unleashed by digital innovation in electricity systems support digital technological development and society-wide digital transformation.

Box 1. Understanding the relationship between electricity and AI

AI is a transformative digital technology that is developing quickly, propelling change across specific industries like energy as well as the wider economy. AI inference and model training primarily takes place in data centres, which are facilities used to house servers, storage systems, networking equipment and associated components to perform computational tasks.

The overall energy impact of AI use depends on several factors: first, the efficiency of the algorithms and hardware that underpin the relevant AI system; second, the extent to which AI is deployed throughout economies and the intensity of that usage; and finally, the specific capabilities of the AI model itself. For example, advanced AI applications, like video generation or agentic tasks, require several orders of magnitude of energy per task than simple text-based generation.

Digital innovation in the wider electricity system is essential to enable the progress of AI. AI is primarily powered by data centres connected to electricity systems, and the global electricity demand of AI-focused data centres grew by 50% in 2025, well outpacing global energy electricity demand growth of 3% in the same year.

Increasing the efficiency of the overall electricity system through digital innovation can help to better accommodate these increases in demand without requiring as much new generation. Similarly, digital innovation in the grid to enable better transmission and distribution can help to connect more data centres, alleviating a key bottleneck for compute going forward. This paper explores AI-driven and other digital innovations in electricity systems which can help support the progress of AI.

Source: IEA (2026[5]) Key questions on AI and Energy, https://www.iea.org/reports/key-questions-on-energy-and-ai; IEA (2025[6]) Energy and AI, https://www.iea.org/reports/energy-and-ai

This paper examines trends in innovative digital applications in the generation, distribution and transmission, use and storage of electricity, or what this paper refers to as “electricity systems”. Increasingly advanced techniques to analyse data captured from connected devices, such as AI, enable a virtuous cycle of monitoring, analysis and efficiency improvements in complex electricity applications. While digital innovative activity in electricity systems remains on an upward trend, funding from the public sector into digital innovation has declined as a share of overall energy public budgets in recent years, potentially threatening the transformative potential of digital technologies for electricity systems. Considering this decline in public investment, this paper then ends with policy considerations to encourage digital innovation in electricity systems.