Urgent Call for Tech Companies: Transparency in Data Centre Energy Usage Amid AI Surge

Experts are urging that tech companies must legally report their energy and water use for data centers. This call comes as artificial intelligence (AI) rapidly expands, raising serious environmental concerns.

A recent report from the National Engineering Policy Centre (NEPC) highlights that AI’s growth is outpacing other energy-consuming systems, which threatens the environment. AI companies are pushing for bigger and faster systems, but this often overlooks how efficiently they use resources.

Prof. Tom Rodden from the University of Nottingham, who helped create the report, warns that this unchecked growth could lead to lasting damage to our planet.

Both Google and Microsoft have noticed significant increases in their data center water use since 2020. Microsoft reported that in 2022, it used 6.4 million cubic meters of water, a jump of 34% from the previous year. Google’s data centers consumed 19.5 million cubic meters in the same year, a 20% increase. Much of this water comes from drinking water supplies, raising further alarms about sustainability.

Currently, there is a lack of precise information on the resources consumed by data centers, making it hard for the government to formulate effective policies to reduce energy usage. To address this, the report emphasizes the need for reliable data to understand the environmental impact of AI.

Rodden insists that gathering trustworthy data is crucial. Once this data is available, we can target efforts to improve efficiency and plan for a greener future in AI.

The report outlines several recommendations for policymakers:

• Expand mandatory reporting of environmental impacts.
• Provide clarity on how AI systems affect the environment.
• Implement sustainability requirements for data centers.
• Reevaluate practices related to data collection and management.

Sarvapali (Gopal) Ramchurn, a professor of artificial intelligence at the University of Southampton, adds that while AI has surged in popularity and usefulness, many people are unaware of its environmental costs. He emphasizes the need for transparency among AI providers. If we can’t measure these impacts, we can’t manage them effectively. The report’s suggestions are intended to inform national discussions on making AI systems more sustainable.

Source link