UN 
(SDGTALKING) – A new report reveals that the growth rate of artificial intelligence technologies is sharply increasing energy demand and deepening data centres’ dependence on fossil fuels. According to experts, this trend in the technology sector threatens global climate goals.
Data Centres’ Energy Hunger is Growing
The training processes of artificial intelligence systems require the processing of billions of parameters. Energy consumption in this process is many times greater than that of traditional computing infrastructure. The report states that the energy demand of AI-based data centres will increase by 160 per cent in the 2024–2025 period. A significant portion of this demand is still met by electricity generated from fossil fuels. Experts point out that despite major technology companies announcing plans to transition to renewable energy, the use of coal and natural gas is increasing in many regions where data centres are located due to energy supply security concerns.
Climate Targets at Risk
To maintain the 1.5°C limit under the Paris Agreement, emissions must be halved by 2030. However, according to the report, the energy intensity of AI-based data infrastructure conflicts with these targets. Researchers emphasise that training a single large language model consumes as much electricity as thousands of households use annually, rapidly increasing the carbon footprint.
Call for Renewable Energy to Technology Companies
United Nations Secretary-General António Guterres recently addressed the technology sector, emphasising that data centres must be powered by renewable energy. ‘Artificial intelligence should serve humanity, not at the expense of the planet,’ said Guterres, noting that accelerating the sector’s energy transition is a ‘moral imperative.’ The report makes a similar call: it recommends raising energy efficiency standards, encouraging waste heat recovery in data centres, and making renewable energy agreements mandatory.
Digitalisation or Carbonisation?
Experts say that if digitalisation in the age of artificial intelligence is not pursued in parallel with sustainability strategies, there is a risk of a ‘carbonised digital future.’ Data centres, cloud computing and artificial intelligence applications are now seen not only as technology investments but also as energy policy issues.
The Transformation Must Begin Now
If the pace of artificial intelligence innovation does not advance in tandem with the energy transition, the sector’s carbon emissions will increase rapidly. The key message highlighted in the report is clear: the future of artificial intelligence is only possible with sustainable energy.
