AIとESG保証の品質：非財務報告の信頼性を高める監査人専門性の役割
AI and the Quality of ESG Assurance: The Role of Auditor Expertise in Increasing the Reliability of Non-Financial Reporting (原題)
Essam AlArbed
🤖 gxceed AI 要約
日本語
米国企業100社・2016〜2025年のパネルデータを用い、AI導入がESG保証品質（ESGQUAL）に与える影響と、公認会計士（CPA）専門性の調整効果を検証した。AI導入とCPA専門性はいずれも保証品質を直接高め、両者の交互作用も有意に正であった。AIは保証の主観性を低減し、その効果はCPA専門性が高いほど増大する。技術補完モデルを支持する結果である。
English
Using a 2016–2025 panel of 100 U.S. firms, this study examines how AI adoption affects ESG assurance quality and whether CPA expertise moderates this link. AI adoption and CPA expertise each directly improve assurance quality, and their interaction is significantly positive. AI also reduces assurance subjectivity, with its effect growing at higher expertise levels, supporting a technology-complementarity model.
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・有報でのサステナビリティ開示が進む日本では、第三者保証の品質確保が実務課題となる。AI活用と会計士専門性の補完関係は、保証業務の設計や監査法人の体制構築に示唆を与える。
In the global GX context
As ISSB/CSRD assurance requirements tighten globally, this paper offers empirical evidence on how AI and auditor expertise jointly raise non-financial reporting reliability. It informs assurance-provider strategy and the debate over technology's role in professional judgment.
👥 読者別の含意
🔬研究者:AIと監査専門性の交互作用がESG保証品質を高める実証的証拠を提供する。
🏢実務担当者:AI導入と会計士関与を組み合わせることで、非財務報告の保証品質と信頼性を高められる。
🏛政策担当者:ESG保証の質を高めるには、AI活用と専門家の判断を両立させる制度設計が有効である。
📄 Abstract（原文）
This study examines how Artificial Intelligence (AI) adoption affects ESG assurance quality (ESGQUAL) and whether Certified Public Accountant (CPA) expertise moderates this relationship. Using a longitudinal panel of 1,000 firm-year observations from 100 U.S. firms spanning 2016 to 2025, the analysis employs firm- and year-fixed-effects regressions with firm-clustered robust standard errors. Results reveal a positive direct association between AI adoption and ESG assurance quality (B = 0.158, p < 0.01) as well as a positive direct effect of CPA expertise (B = 0.176, p < 0.01). Crucially, the interaction term (AI × CPA) is positive and significant (B = 0.089, p < 0.01), confirming that CPA expertise strengthens the beneficial impact of technology. The full model accounts for 48% of the variance in assurance quality ( =0.480). Further tests show AI adoption significantly reduces ESG assurance subjectivity (B = -0.143, p < 0.01). Marginal effects analysis indicates that AI's impact is statistically insignificant at low CPA expertise (B = 0.042), but increases to 0.161 at mean levels and 0.284 at high levels of expertise (p < 0.01). Robustness checks using alternative AI metrics, investment proxies, and lagged specifications confirm these results. The evidence supports a technology-complementarity model where AI enhances data scale and analytical consistency, while CPAs retain core responsibility for professional judgment, ethical oversight, and contextual interpretation.
🔗 Provenance — このレコードを発見したソース
- openalex https://doi.org/10.13189/ujaf.2026.140301first seen 2026-10-13 04:36:47
🔔 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。