Sustainability accounting practices noticeably raise ESG reporting quality, particularly for environmental and governance measures; AI and cloud tools bolster data reliability and frequency, but uptake is concentrated among larger firms and advanced economies, risking an informational divide.

Main Finding

Sustainability accounting practices (SAP) are central to improving the quality of ESG reporting. Across 156 peer‑reviewed studies (2001–H1 2026), SAP were shown to strengthen measurement systems and internal controls and to promote uptake of recognised reporting frameworks (GRI, SASB, ISSB). These effects are strongest for environmental and governance dimensions; social reporting lags because of measurement challenges. Digital technologies — notably AI, cloud platforms, and automation — have become important enablers of data reliability, timeliness, and efficiency, but adoption is uneven, with smaller firms and organisations in emerging economies often falling behind. The literature stresses the combined influence of accounting practice, regulation, corporate governance, and technology on producing more credible and useful sustainability information and highlights the need to reduce symbolic disclosure and improve real‑world impacts of ESG reporting.

Key Points

• Scope and approach
  • Systematic literature synthesis of 156 peer‑reviewed articles spanning 2001 to mid‑2026, analysed through the TCCM (Theory, Context, Characteristics, Methodology) framework.
• Role of SAP
  • SAP strengthen measurement systems, internal controls, and data governance tied to ESG reporting quality.
  • SAP facilitate adoption of established frameworks (GRI, SASB, ISSB), improving comparability and perceived credibility.
• Dimension-specific outcomes
  • Environmental and governance reporting show clearer improvements in measurement and auditability.
  • Social reporting remains weaker due to definitional ambiguity, measurement difficulties, and dispersed data sources.
• Technology and digitisation
  • Over the past decade, AI, cloud services, and automated tools have materially improved data capture, validation, aggregation, and near‑real‑time reporting.
  • Technology reduces transaction costs of reporting but also introduces new risks (data bias, model opacity) and requires digital capability.
• Adoption gaps
  • Smaller firms and organisations in emerging economies have lower uptake of advanced SAP and digital tools, widening informational inequality.
• Persistent problems
  • Symbolic or "greenwashing" disclosures remain an issue; stronger assurance, better metrics, and incentive alignment are needed.
• Research and policy needs
  • Better measurement for social outcomes, study of technology diffusion and equity, improved assurance mechanisms for data and algorithmic outputs.

Data & Methods

• Evidence base: 156 peer‑reviewed articles published 2001–H1 2026.
• Analytical lens: TCCM framework used to structure synthesis — mapping theoretical perspectives, contextual settings (jurisdictions, firm types), characteristics of SAP and reporting outcomes, and the methodologies employed across studies.
• Typical methods in the reviewed literature:
  • Empirical quantitative analyses (archival, regression, event studies)
  • Qualitative case studies and interviews (especially on organisational practices and implementation barriers)
  • Mixed‑methods and conceptual/theoretical papers (framework development)
  • Emerging use of bibliometric and meta‑analytic approaches in later years
• Limitations noted in the review:
  • Heterogeneity in definitions and measures of ESG quality across studies
  • Rapid evolution of digital tools (particularly AI) means recent developments may outpace peer‑reviewed coverage
  • Potential publication bias toward studies in higher‑income jurisdictions and English‑language outlets

Implications for AI Economics

• Productivity and cost structure
  • AI and automation lower the marginal cost of collecting, validating, and reporting ESG data, changing firms’ cost structures for sustainability disclosure. This can enable more frequent and granular reporting, affecting firm disclosure incentives and market reactions.
• Information markets and asset pricing
  • Higher‑quality, higher‑frequency ESG data produced with AI can reduce information frictions for investors, potentially altering pricing of sustainability risks and returns to “green” investments. Economists should model how improved ESG signal quality affects capital allocation and cross‑sectional returns.
• Distributional effects and digital divide
  • Uneven adoption (small firms, emerging markets) could exacerbate informational asymmetries, shifting investment and regulatory attention toward better‑equipped firms or jurisdictions. Policy interventions or subsidised digital infrastructure may be needed to avoid concentration of ESG‑attractive capital.
• Incentives, governance, and strategic behaviour
  • AI can both reduce symbolic disclosure (by improving measurability) and enable sophisticated greenwashing (through selective data pipelines or tailored narratives). Understanding strategic interactions between firms, auditors, and regulators in presence of AI is a priority.
• Audit, assurance, and algorithmic transparency
  • New markets for algorithmic assurance and model auditing will emerge. AI economics research should investigate optimal regulation and market design for auditing AI‑generated ESG outputs, including standards for model explainability, provenance, and data lineage.
• Measurement of social outcomes
  • Social dimensions remain hard to quantify; AI can help (natural language processing, image/video analysis, alternative data) but risks biased or proxy‑driven measures. Economists should evaluate validity, bias, and welfare implications of AI‑based social metrics.
• Research directions
  • Cost–benefit analyses of AI deployment in ESG reporting across firm sizes and jurisdictions.
  • Causal studies on how AI‑enabled reporting affects investment flows, credit conditions, and firm behaviour.
  • Market design work on certification, third‑party assurance, and liability rules for AI‑generated sustainability disclosures.
  • Studies on diffusion barriers, digital infrastructure policy, and targeted support to reduce adoption gaps in emerging markets.
  • Analysis of equilibrium effects if ESG data quality increases substantially (e.g., changes in risk premia, herd behaviour, or regulatory arbitrage).
• Risks to monitor
  • Algorithmic bias, privacy harms from alternative data, model opacity enabling tokenistic disclosures, and concentration of analytics capability among large vendors — all have economic consequences that warrant regulatory and academic attention.

If you want, I can convert these implications into testable research questions or a prioritized agenda for AI‑economics research on ESG reporting.