|
Work item:
|
L.EnvAI (ex L.ClimAI)
|
|
Subject/title:
|
Guidelines for Assessing the Environmental Impact of Artificial Intelligence systems
|
|
Status:
|
Under study [Issued from previous study period]
|
|
Approval process:
|
AAP
|
|
Type of work item:
|
Recommendation
|
|
Version:
|
New
|
|
Equivalent number:
|
-
|
|
Timing:
|
2026-06 (Medium priority)
|
|
Liaison:
|
CEN/CENELEC JTC21, IEEE P7100 and ISO/IEC SC42
|
|
Supporting members:
|
China Unicom, ChinaTelecom, State Grid Corporation of China, ZTE Corporation (China), Nokia corporation, Orange, Colombia, ARCEP
|
|
Summary:
|
As the demand for artificial intelligence technology increases, so does the environmental pressure it imposes. The environmental impact associated with AI, driven by significant energy costs, has become a critical issue that must be addressed for the future development of AI.
This Recommendation provides a holistic framework for evaluating the environmental impact of AI systems, covering direct and indirect impacts, assessment, and mitigation strategies. The methods in L.1410 and L.1480 provide the baseline for this framework. This Recommendation is built on those methods and provides additional guidance to the practitioner how to apply the methods for AI systems.
For minimizing and mitigating the environmental impact of AI systems, it is important to take a life cycle approach understanding the whole life cycle of AI systems. For this, Life Cycle Assessment (LCA) is the most widely used and best available methodology for assessing the environmental impact, including GHG emissions. Therefore, ITU-T L.1410/EN 203 199 has been takes as the baseline for this framework.
This Recommendation can further help relevant parties ensure that current actions are effective in addressing the evolving climate change over the long term. This Recommendation provides a focused framework for evaluating whether it is environmentally beneficial, from environmental impact perspective, to use AI technology compared to not using AI or comparing environmental impact of two AI systems This assessment takes into account the full life cycle of AI systems, including the use of the AI system. It can help AI users to choose a system or solution that will cause less impact to the environment. The methodology followed in this document does not presume a result of the assessment of the use of AI, which may be either contribute to or undermine global sustainability.
|
|
Comment:
|
-
|
|
Reference(s):
|
|
|
Historic references:
|
|
Contact(s):
|
|
| ITU-T A.5 justification(s): |
|
|
|
|
First registration in the WP:
2024-07-03 11:52:59
|
|
Last update:
2025-07-02 09:24:53
|
