Advanced economies face AI-Energy trade-off

The global race to lead in artificial intelligence (AI) has intensified, with advanced economies pouring billions into AI research, infrastructure, and private investment. Still, the surge in digital transformation is unfolding alongside mounting climate commitments and growing scrutiny over the energy consumption of data-driven technologies. The debate is no longer whether AI boosts innovation, but whether it can do so without undermining sustainability.

Addressing this tension, the study, “Can Artificial Intelligence Drive Sustainable Growth? Empirical Evidence on the AI–Energy–Growth Nexus in Advanced Economies,” published in Sustainability, evaluates whether AI investment contributes to short-run economic growth and how renewable energy capacity shapes that outcome.

AI investment alone does not guarantee short-run growth

The study applies four econometric models using a Fixed Effects estimator with Driscoll–Kraay robust standard errors to account for cross-country differences and statistical dependencies. Two sets of models analyze annual data from 2010 to 2025, while additional specifications focus on a shorter 2017 to 2025 period to capture more recent AI investment trends.

The analysis revolves around a key question: does AI investment directly boost short-run economic growth? The findings challenge the assumption that more AI spending automatically translates into higher GDP per capita. Across the tested models, AI investment alone does not produce a statistically significant positive effect on short-run economic growth.

Early-stage adoption of major technologies often involves adjustment costs, learning effects, and transitional inefficiencies. Firms investing in AI must integrate new systems, retrain workers, restructure workflows, and absorb capital costs before productivity gains fully materialize. These transitional dynamics can temporarily dampen growth effects in the short run.

The findings align with both neoclassical growth theory and endogenous growth theory perspectives. From a neoclassical angle, AI can be modeled as a technological shock that enhances total factor productivity and promotes capital deepening. However, these effects are not instantaneous. From an endogenous growth standpoint, AI functions as knowledge capital that accelerates innovation and spillovers, but again, structural transformation takes time.

In practical terms, the study suggests that policymakers should temper expectations that AI spending alone will deliver immediate macroeconomic growth dividends. The growth payoff may depend on complementary conditions, particularly in the energy domain.

Renewable energy as the missing link in the AI–growth nexus

When AI investment is interacted with the share of renewable electricity capacity, the relationship shifts. AI investment combined with higher renewable energy capacity produces a positive and statistically significant effect on short-run economic growth.

This interaction effect reframes the debate. The issue is not simply whether AI promotes growth, but whether AI is embedded within an energy system capable of supporting digital expansion sustainably. In economies where renewable capacity is stronger, AI-driven productivity gains are amplified rather than offset by rising energy costs or environmental constraints.

The study’s conceptual framework emphasizes two opposing mechanisms through which AI affects energy demand. On one side is the expansion effect: data centers, cloud computing infrastructure, automation systems, and high-performance computing increase electricity consumption. On the other side is the efficiency effect: AI can optimize production processes, improve energy management, reduce waste, and enhance resource allocation.

Renewable energy capacity acts as a moderating factor that shapes the balance between these mechanisms. In systems heavily reliant on fossil fuels, increased AI-driven energy demand can weaken sustainability outcomes and dilute economic gains. In contrast, where renewable infrastructure is well developed, AI expansion can coexist with environmental targets, allowing growth and sustainability to reinforce one another.

This conditional relationship provides one of the study’s strongest policy implications. AI and green energy transitions should not be treated as separate agendas. Instead, they must advance in tandem.

Energy demand, environmental pressures, and structural controls

The study also investigates how AI influences energy demand directly. Using per capita energy use as a dependent variable in complementary models, the analysis captures both nonlinearities and interaction effects between AI indicators and renewable energy capacity.

The results acknowledge that AI expansion can increase energy demand in the short run due to digital infrastructure buildout. However, by incorporating renewable capacity and macroeconomic controls, the models account for structural differences among countries, including trade openness, financial development, carbon emissions, energy prices, gross fixed capital formation, and research and development expenditure.

These control variables serve a critical role in isolating AI’s impact. For instance, R&D expenditure reflects innovation capacity and technological readiness. Gross fixed capital formation captures investment intensity, including spending on energy-efficient technologies. Energy price levels influence incentives for conservation and efficiency improvements. Carbon emissions per capita proxy differences in production structure and energy intensity.

By embedding AI within this broader macro-energy framework, the study avoids simplistic conclusions. Instead, it presents AI as part of a complex system in which growth, energy consumption, environmental constraints, and institutional factors interact dynamically.

The time horizon selection further strengthens the analysis. By covering 2010 to 2025 and incorporating a shorter window beginning in 2017, the study captures both early digital transformation trends and the recent surge in AI investment measured through proxies such as ICT goods and services exports and private AI investment indices.

Implications for advanced economies

The sample includes the G7 economies along with China and South Korea, representing technologically advanced nations at the forefront of AI investment and energy transition debates. These countries face similar structural challenges: balancing competitiveness in digital technologies with climate commitments and energy security concerns.

The findings suggest that AI can contribute to sustainable growth, but only under conditions where renewable energy capacity expands alongside digital infrastructure. Without sufficient green energy support, AI’s energy-intensive phase may offset productivity gains, at least in the short term.

Policymakers must integrate digital strategies with climate and energy policy. Investments in AI research, digital infrastructure, and automation should be matched by parallel investments in renewable electricity generation, grid modernization, and energy efficiency systems.

For businesses, the study recommends that ccorporate AI strategies that rely heavily on data processing and cloud infrastructure must account for energy sourcing. Firms operating in regions with stronger renewable penetration may enjoy more sustainable growth outcomes than those in fossil-fuel-dependent environments.

The study also contributes to the broader debate surrounding the Environmental Kuznets Curve hypothesis, which suggests that economic growth initially increases environmental degradation before eventually reducing it.