Generalized green and red concepts for artificial intelligence, big data, and information and communication technologies: Multidimensional framework and countermeasures

A research study introduces Generalized Green AI (GGAI) and Generalized Red AI (GRAI) frameworks to capture the expanding environmental footprint of artificial intelligence, big data, and ICT. Highlighting that U.S. data center electricity consumption could reach up to 12 percent by 2028 and AI water demand could hit 6 billion cubic meters annually by 2027, the authors argue current green AI discussions overlook broader lifecycle, water, and material externalities. Note that the source text provided is an incomplete abstract.

The rapid expansion of computing infrastructure is shifting AI governance from simple carbon management to complex, multi-resource ecological risk management. As electricity grids and local water tables face measurable strain from data operations, technological scaling faces hard physical constraints. Developing comprehensive frameworks to account for these trade-offs is critical to sustaining long-term digital transformation and avoiding systemic resource bottlenecks.

U.S. data center electricity consumption is projected to surge from 4.4 percent in 2023 to between 6.7 percent and 12 percent by 2028.

Forward consideration, not a verified fact.