AI Literacy: The strategic risk energy leaders are underestimating

Reference Paper: International Energy Agency, World Energy Outlook Special Report “Energy & AI”

AI is simultaneously becoming one of the fastest-growing sources of electricity demand and one of the most powerful optimisation tools in the grid. Yet AI-related skills remain significantly less prevalent in the energy sector than in many other industries. That gap is becoming strategic.

Electricity demand from data centres reached around 415 TWh in 2024 and is projected to rise to roughly 945 TWh by 2030 - more than doubling within six years. In advanced economies, data centres are expected to account for more than 20% of electricity demand growth this decade. For utilities that have operated in relatively flat load environments, this represents a structural shift rather than a temporary spike.

At the same time, AI is increasingly embedded inside the energy system itself. AI-based fault detection can reduce outage durations by 30-50%. Advanced optimisation tools could unlock up to 175 GW of transmission capacity globally without building new lines. What was once manual and experience-led is becoming automated and model-driven.

The sector is therefore undergoing a dual transition: serving AI-driven demand growth while operating through AI-enabled systems. The risk is not that AI is advancing too quickly. It is that leadership understanding is not advancing quickly enough.

When AI influences load forecasting, congestion management, renewable integration and predictive maintenance, executives are no longer overseeing purely human processes. They are governing human-machine systems. That changes the nature of oversight and accountability.

AI literacy at leadership level does not mean technical expertise. It means understanding where AI is embedded, how models are validated, what assumptions underpin outputs and where accountability sits. Boards should be able to ask clear questions: How was this model tested? How frequently is it recalibrated? What are the failure scenarios? Who owns the risk?

Without that literacy, three risks emerge.

2) Governance blind spots 

As automated systems influence operational decisions, accountability does not disappear. If an AI-assisted forecasting model misjudges peak demand or a congestion tool allocates capacity incorrectly, responsibility remains with leadership. Governance frameworks must evolve alongside deployment.

3) Resilience exposure 

Cyberattacks on energy utilities have increased sharply in recent years. AI strengthens defensive capability, but it also increases system complexity and digital interdependence. Energy security is expanding beyond fuel diversity and physical infrastructure to include data integrity, supply chain dependencies and algorithmic reliability. Concentrated data centre growth adds another layer of infrastructure risk, particularly where load clusters intensify grid pressure.

None of this suggests that AI is a threat. The upside is substantial. Widespread adoption of existing AI-led efficiency applications could reduce emissions equivalent to around 5% of energy-related emissions by 2035. AI can enhance reliability, improve asset utilisation, accelerate renewable integration and unlock capacity without costly expansion.

But capturing that upside requires informed leadership.

For decades, utilities operated within relatively stable demand environments and clearly defined operational structures. AI is reshaping both. It is accelerating load growth in some regions while simultaneously transforming internal decision-making processes. The constraint is no longer computing power. It is executive fluency.

Companies that treat AI as a narrow digital initiative risk unclear accountability and missed strategic implications. AI is already reshaping demand patterns, operational decision-making and infrastructure planning. Without a clear understanding of how these systems work and how they can disrupt existing business models, leadership teams risk governing technology they do not fully understand.

Utilities that elevate AI literacy to the executive and board level will be far better positioned to align infrastructure planning, resilience strategy and decarbonisation objectives.