Making AI sustainable by design is key to greener future - World Economic Forum (www.weforum.org)

The World Economic Forum has underscored the urgent need to embed sustainability into the design, measurement, and governance of artificial intelligence (AI) technologies. As AI becomes increasingly integral to achieving global sustainability goals, its energy and resource consumption pose significant environmental risks. By 2028, AI could account for over half of data center energy use, equating to the yearly consumption of 22% of US households. This paradox highlights the necessity for a shift in focus from mere accuracy to efficient and responsible resource use in AI development processes. To address these challenges, experts propose frameworks that prioritize ecological performance, such as the AI Energy Score, which encourages energy-efficient practices across the industry. Practical steps include designing energy-efficient models, optimizing data center operations, and extending hardware lifecycles. The call for collective action appears vital, as the complexity of AI's environmental impact necessitates collaboration among energy providers, developers, and regulatory bodies. Adopting standardized methods for measuring AI's ecological footprint will enable organizations to identify key areas for improvement, ultimately fostering an AI ecosystem that aligns performance with environmental stewardship and contributes to a more sustainable future.