From Digital Claims to Verifiable Outcomes: A Deployment-Oriented AI and IoT Disclosure Index Linked to Emissions Intensity and Safety Operations
Elham Jahanbakhsh
🤖 gxceed AI 要約
日本語
本研究は、米国企業の年次報告書やサステナビリティ報告書からAI・IoT導入開示指数（AIDI）を開発し、EPA温室効果ガス報告プログラムの排出原単位や労働安全指標との関連を検証。固定効果モデルにより限定的な関連性を確認し、開示に基づく説明責任の限界を指摘する。
English
This study develops a deployment-oriented AI and IoT disclosure index (AIDI) from U.S. annual and sustainability reports (2018-2024) and tests its association with EPA GHG emissions intensity and workplace injury rates. Using fixed-effects models, it finds limited within-firm links, highlighting constraints on disclosure-based accountability and offering an open tool for assessing digital claims.
Unofficial AI-generated summary based on the public title and abstract. Not an official translation.
📝 gxceed 編集解説 — Why this matters
日本のGX文脈において
米国企業を対象とするが、AI/IoT開示と実排出削減の関連性を問う手法は、日本企業のデジタル化開示がGXにどの程度貢献しているかを評価する枠組みとして参考になる。ただし、日本の開示制度（有報、統合報告書）への直接応用には調整が必要。
In the global GX context
This paper contributes to global GX disclosure scholarship by moving beyond broad ESG scores to a granular, deployment-oriented index of AI/IoT claims linked to actual emissions and safety outcomes. It provides a replicable methodology for regulators and standard-setters (e.g., ISSB, SEC) seeking to verify digitalization's role in decarbonization and offers a cautionary note on over-reliance on disclosure.
👥 読者別の含意
🔬研究者:Provides a text-as-data method for constructing and validating technology disclosure indices linked to environmental and safety outcomes, with open tools for replication.
🏢実務担当者:Offers a framework to critically assess the credibility of firms' AI/IoT claims in sustainability reports and to focus on verifiable operational improvements.
🏛政策担当者:Highlights the limitations of disclosure-based accountability for digitalization claims and suggests the need for facility-level verification mechanisms.
📄 Abstract（原文）
Abstract Amid climate regulation and the low-carbon transition, firms increasingly disclose operational digitalization for cleaner production, yet existing studies relying on broad digital proxies or environmental, social, and governance (ESG) scores limit inference about what these claims signal for environmental management and workplace safety. A deployment-oriented text-as-data index of disclosed Artificial Intelligence (AI) and Internet of Things (IoT) deployment (AIDI) is developed and validated from U.S. annual filings and sustainability reports (2018–2024; 8,605 firms; 36,514 firm-years) using human coding, innovation benchmarks, and a placebo “digital buzz” measure. The index is then linked to operational outcomes—the U.S. EPA Greenhouse Gas Reporting Program (GHGRP) emissions intensity and workplace injury rates (total recordable incident rate; TRIR)—and tested in fixed-effects models with multiplicity adjustment, yielding limited within-firm links; facility- and process-level deployment constrains firm-year verification. The study further offers an open tool for assessing AI/IoT disclosures, identifies limits to disclosure-based accountability, and shows a robust safety–downside risk link, informing sustainability assessment and stakeholder assurance.
🔗 Provenance — このレコードを発見したソース
- openaire https://doi.org/10.21203/rs.3.rs-8960368/v1first seen 2026-05-14 22:34:09
🔔 こうした論文の新着を逃したくない方は キーワードアラート に登録（無料・3キーワードまで）。
gxceed は公開メタデータに基づく研究支援データセットです。要約・翻訳・解説は AI 支援で生成されています。 最終的な解釈・検証は利用者が原典資料に基づいて行うことを前提とします。