AI drives up CO2 emissions

The use of generative artificial intelligence (AI) will drastically increase the CO₂ emissions of data centers by the end of this decade. Analysts at the investment bank Morgan Stanley have calculated that data centers will emit around 2.5 billion tons of greenhouse gases worldwide by 2030 – three times more than if there were no demand for AI.

High electricity costs and emissions

A large proportion of these emissions are caused by the enormous energy requirements of data centers. Around 60% of the emissions come from the electricity consumption of the facilities, while the rest is caused by the production of building materials and the construction of the buildings.

_{CO2 reduction} solutions: a market worth billions

However, rising emissions are also opening up opportunities for companies specializing in CO₂ reduction. According to Morgan Stanley , the market for decarbonization solutions, including carbon capture, utilization and storage (CCUS), could require investments of up to USD 45 billion by 2030. The analysts see great potential for these technologies in the USA in particular.

Hyperscalers as a driver of demand

Large tech companies such as Microsoft, Google and Apple are driving the development of data centers as they need more and more capacity for AI applications. These so-called hyperscalers will have to save 25 megatons of CO₂ by 2030 alone, which is 20 times more than their current levels.

Conclusion: Data centers in transition

The demand for data centers will continue to rise sharply in the coming years, particularly due to the growing use of AI. At the same time, the pressure to reduce CO₂ emissions will increase. Technologies such as carbon capture and reforestation projects could make a decisive contribution to achieving the climate targets by 2030.