Dual impact of AI on renewable transition: Barrier or enabler

The digital revolution is a double-edged sword for the energy transition. Without supportive policies, AI and internet expansion can increase energy demand and slow decarbonization. However, when embedded in knowledge-driven, efficiency-oriented systems powered by renewable energy, digital tools become powerful enablers of low-carbon growth.

Artificial intelligence (AI) and digital expansion can either advance or slow energy transition in emerging economies, depending on how they are deployed and supported by knowledge systems and infrastructure, states a new study published in Sustainability.

Titled “Bridging the Digital–Energy Divide: Artificial Intelligence, Internet Connectivity, and Knowledge Management”, the paper examines the impact of AI, internet connectivity (INTC), and knowledge management (KM) on explicit energy transition (EET), the visible shift to renewable energy sources, and implicit energy transition (IET), efficiency gains that reduce energy intensity, across the BRICS economies from 2000 to 2022. Using two-way fixed-effects models with advanced econometric tests, the authors assess both direct and indirect relationships among digital drivers and their role in advancing low-carbon development.

How AI affects the transition from fossil fuels

The study reports that AI currently exerts a negative direct effect on IET, largely due to the energy consumption associated with expanding digital infrastructure such as data centers and computational processing. This finding challenges the common assumption that AI always supports decarbonization.

However, when AI is integrated into systems that improve energy efficiency, such as predictive demand forecasting, grid optimization, and renewable energy integration, the indirect effect of AI becomes positive for EET. The authors describe this as a dual effect: while digitalization initially raises energy demand, AI-driven operational improvements can mediate these costs and drive explicit adoption of renewables.

The research further shows that economic growth in BRICS countries continues to correlate negatively with both IET and EET, indicating a strong reliance on fossil-fuel-based industrial expansion. In contrast, financial development and trade openness contribute positively to both forms of energy transition, reflecting the importance of capital access and technology spillovers for green innovation. Education, however, remains negatively associated with transition outcomes, suggesting that current curricula and skills frameworks are not yet aligned with sustainable energy priorities.

Role of knowledge management and internet connectivity

According to the paper, knowledge management (KM) plays a complex role. As a moderator, KM shows mixed effects, appearing negative for IET in baseline models due to uneven deployment across sectors, but positive for EET in nonlinear analysis, suggesting that better-structured knowledge platforms can enhance the capacity to adopt renewable technologies. The findings indicate an inverted-U relationship between KM and IET, where knowledge resources initially improve efficiency but yield diminishing returns if not effectively integrated into practice.

Similarly, internet connectivity (INTC) exhibits an inverted-U effect on both IET and EET. Moderate levels of connectivity boost energy transition by improving access to information and enabling smart energy management tools. Yet excessive digital expansion, especially when powered by fossil-based grids, can increase overall energy demand and undermine gains in efficiency.

These patterns underline the importance of policy alignment and infrastructural readiness. Investment in digital networks must go hand in hand with renewable energy expansion and targeted knowledge-sharing platforms to ensure that the benefits of AI and connectivity are not offset by increased energy use.

Policy directions for bridging the digital–energy divide

The energy transition in emerging economies cannot rely on technology alone. Strategic governance and policy reforms are critical for leveraging digital tools to accelerate decarbonization. The study suggests four priority areas:

• Align AI development with clean power by incentivizing data centers to adopt renewable energy and focusing AI applications on demand management, storage optimization, and forecasting for variable renewables.

• Reform knowledge systems by modernizing educational curricula, expanding workforce training in digital energy skills, and building integrated knowledge platforms that link stakeholders across industries.

• Optimize connectivity growth by expanding broadband and smart infrastructure tied to energy-saving applications, while improving the energy efficiency of networks and devices to prevent rebound effects.

• Expand green finance and trade openness by scaling green bonds, sustainable credit lines, and standards that encourage faster adoption of renewable energy technologies.

The authors note that bridging the digital–energy divide requires both structural investment and institutional innovation. Financial and trade liberalization can accelerate the adoption of green technologies, but policy interventions are essential to ensure that growth in AI and connectivity does not drive up emissions.

A roadmap for emerging economies

The digital revolution is a double-edged sword for the energy transition. Without supportive policies, AI and internet expansion can increase energy demand and slow decarbonization. However, when embedded in knowledge-driven, efficiency-oriented systems powered by renewable energy, digital tools become powerful enablers of low-carbon growth.

The research provides policymakers in BRICS and similar economies with insights on where interventions are most needed. It highlights that financial development and open trade remain reliable accelerators, but education and workforce development require urgent reorientation to meet the demands of the green economy.

The authors also acknowledge the study’s limitations, noting that national-level analysis cannot capture sectoral variations and that the proxies used for AI, KM, and INTC may not reflect all qualitative aspects of these drivers. They call for future research at firm and sector levels and for broader comparative studies across developing and developed economies.