Signs of AI Writing in Nearly 40% of Sustainability Reports Studied

Almost 40% of the sustainability reports in our sample of S&P 500 companies show signs of AI-assisted writing in their opening leadership messages. Our analysis classified 89 out of 224 opening messages from reports tagged 2025 as “AI” or “Mixed”, whereas none of the 60 messages from 2016–2021 received either overall classification. These results provide initial evidence that generative AI is becoming part of sustainability reporting.

Generative AI in sustainability reporting

Generative AI can help companies draft, revise, and present their sustainability reports. Academic research points to potential benefits but also warns that persuasive language may enable selective or misleading communication.¹ Reviewing published reports allows us to see if companies are currently employing these tools in their writing.

We analyze reports from a sample of S&P 500 companies, using the opening CEO or Chair letter as an indicator of AI-assisted writing. This letter outlines the company's priorities and its interpretation of performance, providing a common section to compare across companies.

Nearly 40% show an AI-writing signal

Of the 224 opening messages analyzed, 42 were classified as “AI” and 47 as “Mixed”, giving a combined share of 39.7%. The remaining 135 were classified as “Human”.

Percentages are rounded independently.

When using the same approach on opening messages from 60 reports from 2016–2021, there were no “AI” or “Mixed” classifications.

Early evidence from published sustainability reports

Our analysis adds evidence from published sustainability reports to a field where researchers are examining both AI adoption and its effects on disclosure.

Blankespoor, deHaan and Li, writing in the Journal of Accounting Research, find detectable AI use across several types of financial disclosure.²

In sustainability reporting, Dimes and colleagues conducted an experiment in which students used ChatGPT to produce CEO-report sections. Participants subsequently rated AI-generated disclosures as more trustworthy than real examples, while the researchers found greater greenwashing in the generated material under their scoring approach.³

What this means for reporting

For companies, AI might alter the process of report production, but the ultimate responsibility for the published message should stay with the business. Claims regarding progress must continue to be supported by evidence, and commitments need to accurately reflect decisions and actions.

For stakeholders, the question is whether the report provides a reliable account of performance. AI-assisted writing alone cannot answer that. The underlying data, the balance of the narrative and the scope of independent assurance each need scrutiny.

References

1. De Villiers, C., Dimes, R., & Molinari, M. (2024). How will AI text generation and processing impact sustainability reporting? Critical analysis, a conceptual framework and avenues for future research. Sustainability Accounting, Management and Policy Journal, 15(1), 96–118.

2. Blankespoor, E., deHaan, E., & Li, Q. (2026). Generative AI in Financial Reporting. Journal of Accounting Research, 64(3), 1189–1232.

3. Dimes, R., de Villiers, C., Denny, P., & Leinonen, J. (2026). The Impact of Generative AI on the Perpetuation and Detection of Greenwashing in Sustainability Reports: Insights From an Experiment. Business Strategy and the Environment. Advance online publication.