ESGパフォーマンスと企業のAI開示：中国上場企業からの証拠
ESG performance and corporate AI disclosure: evidence from Chinese listed firms (原題)
Jaehyung Bark, Xiaoying Du
🤖 gxceed AI 要約
日本語
中国A株上場企業の2020〜2023年の12,814社・年データを用い、ESG報告書テキストからAI戦略の明示とAIガバナンス開示の2指標を構築。ESG性能、特にガバナンス柱が両開示と有意に正の関連を示し、ラグ・PSM・Oster boundsでも頑健。AI開示が既存のサステナビリティ報告インフラに埋め込まれていることを示し、2022〜2023年に関連が顕著に強まる。
English
Using 12,814 Chinese A-share firm-year observations (2020–2023), the study builds two text-based indicators of AI strategy and AI governance disclosure from ESG reports. Overall ESG performance—especially the governance pillar—is positively associated with both disclosure outcomes, robust to lags, PSM and Oster bounds. AI disclosure appears embedded in firms' broader sustainability reporting infrastructure, with the link strengthening markedly in 2022–2023.
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
AIガバナンス開示をESG報告の一部として捉える視点は、SSBJ基準や有報でのAI・知財開示を検討する日本企業・規制当局にとって示唆が大きい。中国の先行事例として、日本版AI開示枠組みの設計議論に参照価値がある。
In the global GX context
As standard setters (ISSB, CSRD, SEC) begin to grapple with AI governance and digital-related disclosure, this paper offers early emerging-market evidence that AI reporting rides on existing ESG infrastructure. It informs the debate on whether AI disclosure should be a standalone requirement or integrated into sustainability reporting.
👥 読者別の含意
🔬研究者:ESG報告書テキストからAI開示を定量化する手法と、ESG性能との関連の時系列変化を提示する実証研究。
🏢実務担当者:AI戦略・ガバナンスの開示をESG報告に統合することで、投資家対応と報告効率を両立できる可能性を示唆。
🏛政策担当者:AI開示を義務化する際、既存のESG報告枠組みを活用する設計が有効である可能性を示す。
📄 Abstract（原文）
This study examines whether corporate ESG performance is associated with firms’ AI-related disclosure. Using 12,814 firm-year observations from Chinese A-share listed firms during 2020–2023, we construct two text-based indicators capturing whether firms articulate an AI strategy and disclose AI governance information in their ESG reports. Logistic regressions with year and industry fixed effects and firm-clustered standard errors show that overall ESG performance – and the governance pillar in particular – is positively and significantly associated with both disclosure outcomes. These associations are robust to one-year lagged specifications, propensity-score matching and Oster coefficient-stability bounds. The findings suggest that AI-related disclosure is embedded in firms’ broader sustainability reporting infrastructure, with positive associations observed for the environmental, social and governance pillars alike. This study contributes to the voluntary disclosure literature by developing firm-level measures of AI disclosure from ESG report text and by providing early evidence on the ESG–AI disclosure link in an emerging market, and by discussing conditional implications for investors, regulators and standard setters concerned with responsible AI reporting. Additional temporal analyses show that the ESG–AI disclosure association is substantially stronger during 2022–2023 than during 2020–2021.
🔗 Provenance — このレコードを発見したソース
- openalex https://doi.org/10.1080/00036846.2026.2737401first seen 2026-09-30 05:01:11
🔔 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。