Tech companies emitted 301 million tonnes of carbon dioxide in 2024, signaling a failure to meet critical climate targets as the global race for artificial intelligence dominance accelerates.
The figure reveals a growing contradiction between the public net-zero pledges of the world’s largest digital firms and the operational realities of scaling generative AI.
Industry data indicates that the sector is struggling to decouple business growth from greenhouse gas emissions, with the energy-intensive nature of large language models (LLMs) offsetting gains made in hardware efficiency.
Major players including Microsoft, Google, and Amazon have previously committed to becoming carbon neutral or carbon negative by 2030. However, the massive infrastructure required to support AI has pushed these goals further out of reach.
Microsoft, for instance, has acknowledged a significant rise in its total emissions since 2020. The company attributed this growth primarily to the construction of new data centres and the procurement of the high-performance chips required for AI training.
Google has reported similar trends, with its total greenhouse gas emissions increasing as it expanded its computing capacity to compete in the AI landscape. This growth is largely driven by Scope 3 emissions, which include the carbon footprint of the entire value chain and the construction of physical facilities.
AI Expansion Drives Surging Energy Demand
The surge in emissions is tied directly to the power requirements of modern graphics processing units (GPUs). The International Energy Agency has warned that data centre electricity consumption could double by 2026 due to the proliferation of AI.
Unlike traditional search queries, generative AI requests require significantly more computational power and electricity per interaction. This demand extends beyond the servers to the massive amounts of water required for cooling these facilities, which often puts a strain on local resources.
For African markets, this trend presents a dual challenge. As the continent pushes for digital transformation, the deployment of local data centres is essential for reducing latency and improving data sovereignty.
However, since many African energy grids remain heavily dependent on fossil fuels, the expansion of digital infrastructure in the region risks increasing the overall carbon footprint of the continent’s tech ecosystem.
Investors are also beginning to scrutinize these discrepancies. Environmental, Social, and Governance (ESG) reporting requirements are becoming more stringent, and the gap between climate rhetoric and actual emission data is creating regulatory risks for tech boards.
To mitigate these increases, some firms are pivoting toward unconventional energy sources. There is a growing movement among Big Tech to invest in small modular reactors (SMRs) and nuclear fusion to provide a constant, carbon-free baseline of power for their data centres.
Microsoft has recently entered discussions regarding the restart of nuclear plants, such as Three Mile Island, to ensure a dedicated power supply that does not rely on the volatile public grid.
Despite these efforts, the current trajectory suggests that digital emissions will continue to climb until a fundamental breakthrough in energy efficiency or carbon capture technology is achieved.
The industry now faces the prospect of revising its 2030 climate targets or investing unprecedented sums into renewable energy generation to offset the appetite of the AI revolution.
Regulatory bodies in the European Union and the United States are expected to introduce more rigorous disclosure rules for AI-related energy consumption in the coming year to ensure companies provide transparent data on their environmental impact.
Explore more Tech stories and analysis from Business Elites Africa.
