情報ギャップの解消：AIのScope 3排出量開示

Unofficial AI-generated summary based on the public title and abstract. Not an official translation.

🔬研究者:AI利用の排出をScope 3に組み込む際の制度設計ギャップと実証的空白を提示する。

🏢実務担当者:AI利用の炭素足跡をScope 3開示に含める準備と、CDP・SBTi対応の先取りに有用。

🏛政策担当者:AI排出に関する開示義務の曖昧さを解消する規制・ガイダンス整備の根拠を提供する。

This Essay proposes straightforward, immediately implementable reforms to private carbon disclosures systems to motivate reductions in carbon emissions from Artificial Intelligence (AI). AI-driven data centers account for a rapidly increasing share of electricity demand, fossilfuel grid build out, and therefore carbon emissions, but federal, state, and local governments have struggled to develop a public governance response that balances the benefits of rapid development of AI with its carbon emissions. In recent years, non-governmental organizations (NGOs) have worked with corporations to fill gaps in public climate governance by developing the Greenhouse Gas (GHG) Protocol, the CDP (formerly Carbon Disclosure Project), the Science-Based Targets initiative (SBTi) and many supply chain contracting initiatives. This Essay argues that corporations and other organizations that use AI are a principal source of revenue for AI firms, and these corporate AI customers can motivate AI firms to reduce carbon emissions before longer-term regulatory reforms occur. Adequate information on corporate AI-related emissions is needed, but uncertainty about the disclosure obligations of AI-using organizations is a barrier, and the Essay finds that none of the 2024 and 2025 sustainability reports of S&P 500 companies discloses the carbon footprint of their AI use. The Essay proposes quickly resolving disclosure ambiguities and increasing the market signal for decarbonization through (1) clarification of how GHG Protocol Scope 3 standards include AI-driven emissions, (2) addition of questions to the CDP annual survey to gather and disclose information about AI usage and emissions by reporting organizations, (3) guidance on the inclusion of AI-driven emissions in the SBTi disclosure, commitment, and implementation process, and (4) expansion of efforts by NGOs to extend AI-related emissions disclosure and reduction requirements through supply chain contracting. The initiatives outlined in this Essay not only can drive corporate AI emissions reductions, but also emissions reductions from other organizational customers of AI firms, including state and local governments, public and private school systems, universities, religious and civic organizations, and civic and cultural organizations.

• openaire https://doi.org/10.2139/ssrn.7484302first seen 2026-11-05 04:24:15