REVIEW EFFECTIVENESS IN GENERATIVE AI-ASSISTED SUSTAINABILITY REPORTING: A CONCEPTUAL ANALYSIS OF DECOUPLING AND ORGANIZATIONAL ACCOUNTABILITY
Keywords:
Generative artificial intelligence, Sustainability reporting, Organizational decoupling, Management control, Accountability mechanisms, Disclosure qualityAbstract
Generative artificial intelligence (GenAI) can support the drafting and revision of sustainability reports, but the existence of human review procedures does not necessarily improve disclosure quality. Drawing on a theory-driven narrative literature review, this paper applies an organizational decoupling perspective to distinguish policy–practice decoupling, in which established review requirements are not implemented, from means–ends decoupling, in which review takes place but does not effectively serve disclosure quality. Building on research on human oversight and greenwashing, the paper examines the organizational processes underlying these two forms of failure. It develops a conceptual model linking evidence traceability, accountability relationships that enable questioning, and the integration of management controls. The theoretical analysis suggests that access to information, the authority to raise questions, and action in response are interdependent. Communications-oriented incentives may influence reviewers’ attention to different kinds of evidence, while feedback across reporting cycles influences whether corrections support organizational learning. Three propositions are developed, together with alternative explanations and boundary conditions. The theoretical contribution is to explain why human involvement can produce different outcomes and to distinguish procedural completion, disclosure quality, and sustainability performance. No new experimental or survey data were collected. The proposed relationships remain to be empirically tested.References
[1] Mustafa F, Smolarski J, Elamer A A. The Convergence of Artificial Intelligence and Sustainability Reporting: A Systematic Review of Applications, Challenges and Future Directions. Business Strategy and the Environment, 2025, 34(8): 9761-9784. DOI: 10.1002/bse.70090.
[2] Mousavian Anaraki S A, Croce D, Basili R. Large language models for sustainability reporting: A systematic review and research agenda. Sustainable Futures, 2025, 10: 101494. DOI: 10.1016/j.sftr.2025.101494.
[3] Dimes R, de Villiers C, Denny P, et al. The Impact of Generative AI on the Perpetuation and Detection of Greenwashing in Sustainability Reports: Insights From an Experiment. Business Strategy and the Environment, 2026. Advance online publication. DOI: 10.1002/bse.70983.
[4] Snyder H. Literature review as a research methodology: An overview and guidelines. Journal of Business Research, 2019, 104: 333-339. DOI: 10.1016/j.jbusres.2019.07.039.
[5] Jaakkola E. Designing conceptual articles: four approaches. AMS Review, 2020, 10(1-2): 18-26. DOI: 10.1007/s13162-020-00161-0.
[6] Michelon G, Pilonato S, Ricceri F. CSR reporting practices and the quality of disclosure: An empirical analysis. Critical Perspectives on Accounting, 2015, 33: 59-78. DOI: 10.1016/j.cpa.2014.10.003.
[7] Cho C H, Laine M, Roberts R W, et al. Organized hypocrisy, organizational façades, and sustainability reporting. Accounting, Organizations and Society, 2015, 40: 78-94. DOI: 10.1016/j.aos.2014.12.003.
[8] Ni J, Bingler J, Colesanti-Senni C, et al. CHATREPORT: Democratizing Sustainability Disclosure Analysis through LLM-based Tools. Proceedings of the 2023 Conference on Empirical Methods in Natural Language Processing: System Demonstrations, 2023: 21-51. DOI: 10.18653/v1/2023.emnlp-demo.3.
[9] Ji Z, Lee N, Frieske R, et al. Survey of Hallucination in Natural Language Generation. ACM Computing Surveys, 2023, 55(12): 1-38. DOI: 10.1145/3571730.
[10] Vaccaro M, Almaatouq A, Malone T. When combinations of humans and AI are useful: A systematic review and meta-analysis. Nature Human Behaviour, 2024, 8(12): 2293-2303. DOI: 10.1038/s41562-024-02024-1.
[11] Raji I D, Smart A, White R N, et al. Closing the AI accountability gap: defining an end-to-end framework for internal algorithmic auditing. Proceedings of the 2020 Conference on Fairness, Accountability, and Transparency, 2020: 33-44. DOI: 10.1145/3351095.3372873.
[12] Autio C, Schwartz R, Dunietz J, et al. Artificial Intelligence Risk Management Framework: Generative Artificial Intelligence Profile (NIST AI 600-1). National Institute of Standards and Technology, 2024. DOI: 10.6028/NIST.AI.600-1.
[13] Gond J P, Grubnic S, Herzig C, et al. Configuring management control systems: Theorizing the integration of strategy and sustainability. Management Accounting Research, 2012, 23(3): 205-223. DOI: 10.1016/j.mar.2012.06.003.
[14] Bromley P, Powell W W. From Smoke and Mirrors to Walking the Talk: Decoupling in the Contemporary World. Academy of Management Annals, 2012, 6(1): 483-530. DOI: 10.5465/19416520.2012.684462.
[15] Bovens M. Analysing and Assessing Accountability: A Conceptual Framework. European Law Journal, 2007, 13(4): 447-468. DOI: 10.1111/j.1468-0386.2007.00378.x.