A new report from the United Nations University Institute for Water, Environment and Health highlights the growing environmental footprint of the global data centre industry as artificial intelligence adoption accelerates. According to the research, data centres consumed 448 TWh of electricity and 4.5 trillion litres of water globally in 2024, with AI workloads accounting for approximately 20% of total power demand. By 2030, electricity consumption is projected to reach 945 TWh – equivalent to the current power usage of Japan – while water consumption could more than double to 9.3 trillion litres. The report also forecasts a significant increase in land use and carbon emissions as operators race to deploy new capacity. While AI has the potential to improve energy efficiency across sectors through optimisation and automation, the findings underscore a critical challenge for the industry: balancing unprecedented digital growth with responsible resource management. For GCC markets, where both AI investment and water scarcity are strategic considerations, the report serves as a timely reminder that future competitiveness will depend not only on scale and speed of deployment, but also on sustainable infrastructure design, renewable energy integration, advanced cooling technologies, and long-term resource planning.
SoftBank Commits €75bn to Position France as a European AI Infrastructure Hub
SoftBank has unveiled plans to invest up to €75 billion ($87.5 billion) in developing 5GW of AI-focused data centre capacity across France, marking one of Europe’s largest AI infrastructure commitments to date. The first phase, representing a €45 billion investment, aims to deliver 3.1GW of capacity by 2031 across sites in Dunkirk, Bosquel, and Bouchain. The project reflects a broader trend of hyperscale AI infrastructure clustering around regions with abundant power availability, supportive industrial policy, and established engineering ecosystems. Strategic partnerships with EDF and Schneider Electric will support both power provision and the development of local manufacturing capabilities for prefabricated data centre infrastructure. France’s access to low-carbon nuclear energy, combined with government-backed acceleration of strategic digital projects, is increasingly positioning the country as a leading European destination for AI infrastructure investment. The announcement also highlights the growing scale of capital required to support next-generation AI workloads, with SoftBank continuing to expand its global AI infrastructure ambitions alongside investments in OpenAI and other large-scale data centre initiatives.
New research from York University highlights the rapid acceleration of Canada’s AI infrastructure market, with announced and proposed data centre projects potentially increasing national capacity from 1.6GW today to more than 13GW in the coming years. The shift is being driven by a new generation of hyperscale facilities designed to support AI training and high-performance computing workloads, with many projects significantly larger than traditional enterprise data centres. Alberta has emerged as a key growth market, attracting investment through available power capacity, competitive energy costs, strong fibre connectivity, and a pro-development policy environment. While provinces such as Ontario, Quebec, and British Columbia continue to host the majority of existing facilities, future growth is increasingly concentrating in regions where large-scale power resources can support AI-driven demand. As the sector expands, industry stakeholders are placing greater emphasis on sustainability, resource efficiency, and community engagement, with operators investing in advanced cooling technologies, closed-loop water systems, and new approaches to energy management. The findings underscore a broader global trend: access to scalable power infrastructure is becoming one of the most important factors shaping the next phase of AI data centre development.
A proposed hyperscale data centre program in the UAE could add between 750MW and 1.28GW of AI computing capacity over the next 10 to 15 years, further reinforcing the country’s position as a regional digital infrastructure hub. The initiative, put forward by Dominican private bank Asprofin in collaboration with IDC Technologies, envisions three large-scale campuses requiring investments of between $5 billion and $12 billion, subject to feasibility studies and final design decisions. A key differentiator of the proposal is its emphasis on modular, prefabricated infrastructure, which the developers believe could significantly reduce capital costs compared to traditional hyperscale deployments. While still at an early stage, the project reflects growing industry interest in more efficient approaches to scaling AI infrastructure as demand for high-performance computing continues to accelerate. The announcement also highlights the UAE’s increasing attractiveness for data centre investment, supported by strong connectivity, established cloud regions, favourable digital transformation policies, and a regulatory focus on energy-efficient and sustainable infrastructure development.