サステナビリティ報告書監査における人工知能の応用：倫理的課題と効率性の狭間で

THE APPLICATION OF ARTIFICIAL INTELLIGENCE IN AUDITING SUSTAINABILITY REPORTS: BETWEEN ETHICAL CHALLENGES AND EFFICIENCY (原題)

Galina Bădicu, Svetlana Mihaila, V. Grosu

PROCEEDINGS OF THE INTERNATIONAL SCIENTIFIC CONFERENCE COMPETITIVENESS AND INNOVATION IN THE KNOWLEDGE ECONOMY📚 査読済 / ジャーナル2026-05-12#AI×ESG経営インパクト: 調達リスク対象セクター: finance

DOI: 10.24818/cike2025.52

原典: https://doi.org/10.24818/cike2025.52

🤖 gxceed AI 要約

日本語

サステナビリティ報告書の保証業務にAIを組み込む際の効率性と倫理的整合性を、実務者サーベイとLLM/NLPによるESG報告書分析で検証した研究。倫理的 safeguard（バイアス検査・透明性・独立性）がステークホルダー信頼を高め、AI支援レビューが時間短縮とグリーンウォッシュ検出力向上を両立することを示す。ISSA 5000・IFRS S1/S2・ESRS・GRI準拠のワークフロー設計が証拠の追跡性とコンプライアンス認識を高めると結論づける。

English

This study examines how AI can improve sustainability-report assurance efficiency while preserving ethical integrity and standards compliance. Using a practitioner survey plus LLM/NLP analysis of ESG reports, it finds ethical safeguards (bias testing, transparency, independence) raise stakeholder trust, AI-assisted review cuts time while maintaining or improving greenwashing detection, and aligning AI workflows with ISSA 5000, IFRS S1/S2, ESRS and GRI improves evidence traceability and perceived compliance.

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

SSBJ基準の確定と保証業務の制度化が進む日本では、AI活用監査の倫理ガバナンスと証拠追跡性の枠組みは、監査法人・企業の保証対応設計に直接示唆を与える。

In the global GX context

As ISSA 5000 and CSRD/ESRS assurance requirements tighten globally, this paper offers an integrated ethics-efficiency-compliance model for AI in sustainability assurance, directly relevant to assurance providers and standard setters shaping AI governance in disclosure infrastructure.

👥 読者別の含意

🔬研究者:AI×保証の倫理・効率・基準準拠を統合した三本柱モデルは、ESG保証研究の実証的枠組みとして参照価値が高い。

🏢実務担当者:AI支援レビューの導入時にバイアス検査・透明性・独立性プロトコルを設計し、ISSA 5000等への整合を図る実務指針として使える。

🏛政策担当者:AI活用保証の倫理ガイドラインと継続的職業訓練の制度化を検討する際の根拠として活用できる。

📄 Abstract（原文）

Artificial intelligence is increasingly embedded in the assurance of sustainability reports, promising efficiency gains yet raising acute ethical concerns. This study examines how AI can improve audit efficiency while preserving ethical integrity, and whether such use remains compliant with emerging assurance and reporting standards. Building on a three-pillar model: Ethics and Independence, Efficiency and Detection, and Governance and Compliance, we derive three hypotheses and evaluate them through a practitioner survey of auditors, accountants and senior managers, complemented by a targeted document analysis using LLM/NLP tools on ESG reports. Results indicate that explicit ethical safeguards (bias testing, transparency and independence protocols) are associated with higher stakeholder trust; moreover, AI-assisted review reduces time while maintaining or improving the detection of exaggeration/greenwashing; finally, aligning AI workflows with ISSA 5000, IFRS S1/S2, ESRS and GRI enhances traceability of evidence and perceived compliance. Accordingly, the study underscores the need for continuous professional training, clear ethical guidelines and governance-by-design for AI in sustainability assurance, and offers an integrated model that connects ethics, efficiency and standard-based compliance.

🔗 Provenance — このレコードを発見したソース

semanticscholar https://doi.org/10.24818/cike2025.52first seen 2026-10-30 05:46:27 · last seen 2026-11-06 05:45:25

🔔 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。