AI can boost efficiency in global electricity systems

Recent advancements in artificial intelligence (AI) and its increased adoption can help accelerate the global energy system’s transformation to net zero whilst meeting the projected 3–4 per cent per year increase in global electricity demand through 2030, a report said on Thursday.

Released by Adnoc, Masdar and Microsoft, the report, entitled ‘Powering Possible: AI and Energy for a Sustainable Future’, highlights how AI is critical for accelerating the net-zero energy transformation by boosting efficiencies, decarbonizing existing systems, and expanding carbon-free energy sources.

The electricity needed to power the data centres critical to AI and other digital services is expected to grow between 8–23 per cent per year through 2026, bringing AI use to 0.24 per cent of global electricity demand. “Despite $250 billion in private investments in AI between 2017 and 2023, only 5 per cent of that funding went to the energy sector,” the report noted.

According to the International Renewable Energy Agency (IRENA), the annual rate of deployment of renewable generation must grow by 16 per cent every year through 2030 to meet the tripling target. “To achieve this target, it is critical to address key challenges such as permitting, grid infrastructure, financing, and supply chain bottlenecks,” the report said.

The report highlights AI’s capacity to reduce methane emissions – a greenhouse gas that is 80 times more effective at trapping heat than carbon dioxide (CO2). Advanced AI tools being developed are anticipated to detect methane leaks with up to 20 per cent greater accuracy compared to legacy technologies. These innovations are expected to play a vital role in helping the energy sector achieve the Global Methane Pledge, which aims for a 30 per cent reduction in methane emissions by 2030.

The report also provides a roadmap to address AI-driven data centres’ rising energy demands, which are projected to nearly double their share of global electricity demand by 2026. While at a global scale, AI’s electricity consumption is relatively small, in some regions, electricity demand for data centres can be a large percentage of total load. For example, according to the IEA, in the European Union, where data centre electricity demand is expected to increase at ~9 per cent per year, demand could exceed 5 per cent of the total regional load by 2026. “Building on the significant progress already underway to bring more carbon-free energy online, the collaboration between the technology and energy sectors is crucial to ensuring that growing electricity demand is met through sustainable and carbon-free energy solutions. Masdar, the UAE’s clean energy champion, highlights within the report the opportunity for AI to transform the global clean energy industry, as the data centers the industry requires will serve as an increasingly important driver of global energy demand, adding that to meet this demand sustainably will require a multifaceted approach to unleash the full potential of AI,” a statement said.