Global energy poverty crisis escalates amid climate and AI shocks

The world is facing a growing crisis: the inability of millions to access or afford essential energy services. A comprehensive review published in Energies under the title “Exploring Energy Poverty: Toward a Comprehensive Predictive Framework” by Takako Mochida, Andrew Chapman, and Benjamin Craig McLellan, outlines the urgent need for a redefined and forward-looking approach to evaluating and predicting energy poverty.

As climate change accelerates and energy demand rises with emerging technologies, the current metrics and policy responses are shown to be deeply inadequate, failing to account for mounting exogenous pressures. The study proposes a multidimensional forecasting model that integrates climate impacts, economic shocks, and tech-driven consumption alongside traditional indicators like income and household characteristics.

How has energy poverty been traditionally measured and why is it no longer enough?

Energy poverty, commonly associated with insufficient access to affordable, reliable, and clean energy, has been conventionally assessed through static indicators such as the 10% income expenditure rule in the UK, or multidimensional indexes like the Multidimensional Energy Poverty Index (MEPI). These frameworks often rely on national household survey data, such as energy costs, income thresholds, or appliance ownership. While these tools provide surface-level insights, the study argues they miss the complex interactions between social, technological, and environmental variables.

Critically, the review shows that traditional energy poverty indicators are tailored to the historical context of the Global North, largely ignoring the heating-centric metrics that fail to capture rising cooling demands in warmer or tropical regions. For instance, while Southern Europe has begun to explore summer energy poverty (SEP), many nations, including developed ones like Japan and Australia, lack coherent evaluation models that incorporate both cooling needs and climate volatility.

Furthermore, these metrics fall short of encompassing household behaviors, mental health vulnerabilities, or socio-demographic trends like aging and migration. In Japan, for example, over 60% of fatal heatstroke victims had access to air conditioning but didn’t use it effectively due to cognitive or economic constraints. Such cases highlight the inadequacy of existing tools to address the multifaceted nature of energy poverty under extreme weather conditions.

What new drivers are emerging, and why do they demand a predictive shift?

The study identifies a critical transformation in the drivers of energy poverty. Traditionally attributed to low income, inefficient housing, and high energy costs, energy poverty is now increasingly shaped by global phenomena including climate change, population aging, geopolitical instability, and the rise of energy-intensive technologies such as AI and data centers.

Climate change is a central concern, with phenomena like heat domes and more frequent extreme weather events exacerbating energy burdens. A single heatwave in the U.S. was projected to cause $500 billion in annual economic damage and over 60,000 deaths by 2050. These impacts highlight energy poverty as not just an economic issue, but a public health and climate justice emergency.

Technological advancement, while often touted as a solution, also plays a contradictory role. AI and cryptocurrency mining are forecast to double global electricity consumption by 2026, potentially pushing communities living near data centers into energy poverty due to localized energy price spikes and supply strains. The demographic transition, particularly aging populations in countries like Japan, further complicates the landscape. Elderly individuals, especially those living alone, have higher energy needs but face mobility and financial barriers that inhibit energy-efficient adaptations.

Adding to this complexity, factors such as air conditioning access, fossil fuel price volatility, economic crises, and the aftereffects of pandemics have become short- to medium-term shocks that are neither integrated nor adequately weighted in current policy frameworks.

How should the future of energy poverty prediction and policy look?

The authors propose a paradigm shift from reactive assessment to proactive forecasting. This entails integrating a dynamic, flexible suite of indicators grouped into three categories: long-term trends (e.g., climate change, tech diffusion), short-term shocks (e.g., economic crises, CPI fluctuations), and dual-impact factors (e.g., occupational exposure, demographic shifts).

Figure-based models in the study map the interactions between these variables, underscoring how the interplay between exogenous shocks and endogenous vulnerabilities creates a cascading effect on household energy security. For example, economic downturns reduce disposable income, which combined with rising temperatures and energy prices, intensifies energy-related stress and health risks.

The framework also emphasizes spatial equity. Energy poverty is increasingly regionalized, affecting rural and climate-vulnerable communities disproportionately. Remote sensing and machine learning tools are highlighted as promising in identifying these high-risk areas. In India, such models achieved over 90% accuracy in predicting energy poverty using weather and socio-economic data. Similar initiatives in the Netherlands and UK have integrated housing quality, welfare access, and even consumer switching behavior into predictive indices.

To operationalize this approach, the study recommends embedding energy poverty mitigation into broader climate and energy transition policies. For instance, prioritizing zero-energy buildings (ZEBs) in vulnerable communities could offer dual benefits of carbon reduction and improved household energy resilience. It also calls for governments to develop regulatory safeguards against energy-intensive infrastructure developments like AI data hubs that could deepen regional inequality.

The study urges global and national stakeholders to recognize that energy poverty is not a static condition but a dynamic, climate-sensitive, and socially contingent phenomenon. Addressing it effectively requires a rethinking of both the indicators we use and the systems we design to respond to them.