The surge in demand for generative artificial intelligence (AI) is poised to significantly escalate greenhouse gas (GHG) emissions from data centres, according to a new Morgan Stanley report.
The study reveals that by 2030, data centres are expected to emit a staggering 2.5 billion tons of GHGs—three times higher than previous projections that did not account for the impact of generative AI.
This alarming forecast underscores the urgent need for the tech industry to address its environmental footprint as it continues to grow.
Generative AI drives emissions surge to 600 million tons
The report highlights a dramatic increase in emissions linked to the rising demand for generative AI.
Emissions from data centers are projected to jump from 200 million tons this year to 600 million tons by 2030.
This surge is primarily driven by the expansion of data centers required to support growing cloud services.
According to Morgan Stanley, 60% of these emissions will come from the energy-intensive operations of data centers, while the remaining 40% will result from the production of construction materials and infrastructure.
Data center expansion threatens net-zero targets
The rapid construction of new data centers presents a significant obstacle to achieving net-zero emissions goals.
Emissions generated during the construction phase, including the production of materials and infrastructure, contribute heavily to the overall carbon footprint.
Google’s recent report of a 48% increase in emissions over the past five years exemplifies the challenges faced in meeting environmental targets.
With the tech sector currently responsible for 40% of annual US emissions, the path to achieving net-zero targets appears increasingly fraught.
Efforts to mitigate the environmental impact of data centers face additional challenges, particularly regarding cooling systems.
While water-cooling systems can reduce energy consumption, they require substantial amounts of water, which poses issues in regions experiencing water scarcity.
This complicates the tech industry’s efforts to meet green goals and exacerbates the environmental impact of the growing AI sector.
Morgan Stanley’s study suggests that meeting necessary carbon capture, utilization, and sequestration (CCUS) standards will demand a substantial investment of $15 billion.
The report also identifies reforestation projects as a potential solution to help achieve future net-zero targets.
These findings highlight the urgent need for innovative approaches and significant financial commitment to address the environmental impact of the rapid development of generative AI.
The post Data center greenhouse gas emissions set to triple by 2030 due to generative AI demand appeared first on Invezz
