Want students to understand annual corporate sustainability reports?

Katie Balaam from the Open University looks at how students can use AI to cut through the noise of corporate sustainability reporting without losing critical thinking or ownership of the final analysis.

Corporate reporting is changing – sustainability reporting goes mainstream

Sustainability reporting has become mainstream with a large and increasing number of firms around the globe now disclosing sustainability information (KPMG, 2024KPMG (2024) The move to mandatory reporting: Survey of Sustainability Reporting 2024. KPMG International. Available at: https://kpmg.com/xx/en/our-insights/esg/the-move-to-mandatory-reporting.html (Accessed: 5 November 2026).). This includes, for example, how a business relies on ecosystem services to function and make money, and how its operations alter the health of the natural world. This has been driven by a massive increase in sustainable investment and its association with higher firm values (KPMG, 2024KPMG (2024) The move to mandatory reporting: Survey of Sustainability Reporting 2024. KPMG International. Available at: https://kpmg.com/xx/en/our-insights/esg/the-move-to-mandatory-reporting.html (Accessed: 5 November 2026).; Cahan et al., 2016Cahan, S.F., De Villiers, C., Jeter, D.C., Naiker, V. and Van Staden, C.J. (2016) Are CSR disclosures value relevant? Cross-country evidence. European Accounting Review, 25(3), pp. 579–611. Available at: https://doi.org/10.1080/09638180.2015.1064009). Reporting has historically been largely voluntary and dominated by larger organisations, but the shift toward mandatory, standardised reporting, such as the EU’s Corporate Sustainability Reporting Directive and the ISSB standards, is expected to improve comparability (De Villiers & van Staden, 2011De Villiers, C. and Van Staden, C.J. (2011) Where firms choose to disclose voluntary environmental information. Journal of Accounting and Public Policy, 30(6), pp. 504–525. Available at: https://doi.org/10.1016/j.jaccpubpol.2011.03.005; KPMG, 2024KPMG (2024) The move to mandatory reporting: Survey of Sustainability Reporting 2024. KPMG International. Available at: https://kpmg.com/xx/en/our-insights/esg/the-move-to-mandatory-reporting.html (Accessed: 5 November 2026).).

Concerns regarding corporate deception are rising, with terms like greenwashing and bluewashing being discussed. While greenwashing refers to the practice of inflating or fabricating a company’s environmental credentials to appear ecologically responsible, bluewashing specifically denotes the deceptive manipulation of a firm’s social, humanitarian, or human rights record. Furthermore, sustainability disclosures are often complex, vague, and difficult for non-professional users to interpret. Broader societal efforts encourage individuals to act responsibly in their consumption and investment decisions, making it essential that sustainability-related information is both accessible and understandable to the public. However, comprehension of these disclosures remains low because of their complexity.

What are the problems with corporate reporting?

There is a gap between the availability of sustainability information and the public’s ability to understand and use it effectively. AI can be used to analyse large quantities of data (De Villiers, Dimes, & Molinari, 2024De Villiers, C., Dimes, R. and 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), pp. 96–118. Available at: https://doi.org/10.1108/SAMPJ-02-2023-0097) and could therefore potentially reduce this gap. However, little is known about their actual use or effectiveness in improving the public’s understanding, trust, or decision-making. Sustainability disclosures are often voluntary and lack standardisation and comparability, making them difficult for users to interpret due to the narrative-heavy content (De Villiers, Rinaldi, & Unerman, 2014De Villiers, C., Rinaldi, L. and Unerman, J. (2014) Integrated Reporting: insights, gaps and an agenda for future research. Accounting, Auditing & Accountability Journal, 27(7), pp. 1042–1067. Available at: https://doi.org/10.1108/AAAJ-06-2014-1736; Lewis & Young, 2019Lewis, C. and Young, S. (2019) Fad or future? Automated analysis of financial text and its implications for corporate reporting. Accounting and Business Research, 49(5), pp. 587–615. Available at: https://doi.org/10.1080/00014788.2019.1611730). Prior research on the digital transformation of corporate reporting shows that digital technologies can improve stakeholder engagement and enhance social responsibility, accountability, and transparency (Lombardi & Secundo, 2021Lombardi, R. and Secundo, G. (2021) The digital transformation of corporate reporting – a systematic literature review and avenues for future research. Meditari Accountancy Research, 29(5), pp. 1179–1208. Available at: https://doi.org/10.1108/MEDAR-04-2020-0870). However, the use of AI in relation to sustainability reporting remains under explored (De Villiers, Dimes, & Molinari, 2024De Villiers, C., Dimes, R. and 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), pp. 96–118. Available at: https://doi.org/10.1108/SAMPJ-02-2023-0097; Moodaley & Telukdarie, 2023Moodaley, W. and Telukdarie, A. (2023) Greenwashing, sustainability reporting, and artificial intelligence: a systematic literature review. Sustainability, 15(2), p. 1481. Available at: https://doi.org/10.3390/su15021481).

Given the growing interest in environmental and social impacts, there is an urgent need to ensure that the public can meaningfully engage with and understand sustainability disclosures. Evidence on the role of AI tools in helping general users understand sustainability content in reports, based on evidence collected in the UK, remains limited. With evidence showing ESG disclosures influence investor decisions (Misiuda & Lachmann, 2022Misiuda, M. and Lachmann, M. (2022) Investors’ perceptions of sustainability reporting – a review of the experimental literature. Sustainability, 14(24), p. 16746. Available at: https://doi.org/10.3390/su142416746), a project was established to investigate how AI tools can help understand sustainability information in annual reports. This pilot investigates how students in the UK engage with sustainability disclosures and whether AI can enhance comprehension and perception, through an online pilot survey of Open University students, ahead of a fuller information experiment representing UK households.

Results from a pilot study

Despite advertising the survey through several module forums and via the Open University Students Association, we did not have a high level of participation and had only 11 respondents. We conducted the survey with these respondents and collected data for the pilot run. The sample was split roughly evenly between men and women, with the majority falling in the 25–34 age category. Education varied, however over half have a master’s or higher qualification. Occupation was varied and ranged between working full time, being a full-time student and two who were unemployed.

During the survey, respondents were asked about their investment habits, and it was found that they have a strong desire to back ethical companies and want AI to help them, however interestingly they rarely check their portfolios and do not fully trust AI for final investment decisions.

The pilot found that 9 of the 11 respondents (82%) agreed that AI makes it easier to process large amounts of data in reporting, and only one did not use AI at all. Seven (64%) reported that they only understand some of the sections of the sustainability report and leaned heavily on AI to break down complex sustainability data but do maintain strong human oversight because they do not fully trust the tools. The reason behind this is that sustainability information is too complex to understand alone, as corporate sustainability reports are filled with dense legal jargon, complicated science metrics, and hundreds of pages of text.

The main issue with using AI tools was trust, and respondents said that they didn’t think AI tools are reliable enough to make final, high-stakes financial choices on their own. In particular, they mentioned that AI can make mistakes. Indeed, 8 of the 11 respondents (73%) disagreed with the statement that they can rely on the information that is generated by AI, with only three (27%) agreeing that the AI tools provide reliable explanations.

With regard to investing in sustainable firms, we did not find one unified attitude within the small sample. Examples for sustainable investment included protecting the planet, ensuring workers are treated fairly, and companies practising good corporate governance. However, around half of the respondents rarely take the time to carry out a review of their current investments and are willing to accept slightly lower cash returns provided they are investing with companies with high sustainability measures.

Our findings suggest that respondents find it difficult to fully understand these disclosures without an outside helper or expert tool to translate them. Respondents use AI to help them grasp the core meaning of a company’s sustainability goals. However, although most respondents use AI regularly for these tasks, our findings show that they do not blindly trust the outputs. Users are aware of AI risks like hallucinations (made-up facts) or data miscalculations, so they double-check the final data and still rely on human judgment for final decisions.

Classroom activity: the AI dilemma

AI is developing at such a speed that failing to incorporate it into our teaching is no longer an option. However, with various platforms now available to students with varying degrees of understanding it is vital that we embed how to use AI efficiently within our teaching. For example, instead of telling your students about the pitfalls of AI, let them experience it for themselves. You can select a real-life sustainability report, or devise a mock one and base the following example around it:

Imagine you and your classmates are university students competing in a high-stakes sustainable finance competition. Your team has made it to the final round, where you must consider whether to invest in a multinational corporation for a £70m green investment portfolio. The judges expect a flawless presentation, but there is a major problem: the deadline is in three hours, and you still need to review the company’s 250-page sustainability report.

You could create groups where one group is able to use AI, and the second group is not allowed to use AI.

Discussion and reflection for students

These questions are designed to make the AI group look closely at their generated summary and question its validity.

For example, ask the team who used AI whether they identified any of the following issues;

1. Your AI summary portrays this company’s environmental data as a flawless global achievement. Did the AI give you a breakdown of where this data comes from, or did it assume the rules apply equally to every region?
2. The AI reported a generalised carbon reduction victory. Did it break down the difference between the company’s direct operational emissions and its supply chain emissions, or did it just give you one net average percentage?
3. Your summary confidently states that the data complies with the disclosure requirements under ISSB. Find the exact sentence in your AI text that proves that? If you look at the raw source text, does it match, or did the AI convert a future plan into a present fact?
4. The AI output is grammatically perfect and sounds incredibly professional, how tempted were you to just accept it as 100% accurate without double-checking the source?
5. When you were reading the text, did any specific words or phrases feel like ‘fluff’ or ‘PR spin’ to you? How did you train your brain to ignore the shiny headlines and find the facts?
6. Did you notice any shifts in tone? For example, did the company sound very confident in the main text but suddenly sound highly technical and cautious when discussing regional liabilities or future compliance?

For the group that didn’t use AI, you could ask them a series of question, for example;

1. You found critical information hidden on page 197 in a footnote. What made you look there? Did the layout of the document itself make you suspicious that something was being buried?
2. How did your eyes track the data? Did you read the report from front to back, or did you purposefully skip the introduction and jump straight to the data tables and disclosures?
3. While the AI group finished in just twenty minutes, you had to manually sift through the pages. Did the extra time you spent give you a deeper, more memorable understanding of this company’s actual business ethics?
4. Now that you know the AI completely missed the regional lawsuits and hallucinated the ISSB compliance, how do you feel about the common narrative that ‘AI is replacing the need for human analysts’?

Then bring the whole group together to discuss the following;

1. Now that you’ve heard what the manual group found buried in the footnotes and raw data, how would you grade your AI’s summary on a scale of A to F for accuracy?
2. The AI group finished in 20 minutes. The manual group took longer but found massive regional liabilities and misreported compliance data. If you were a regulator or an investigative journalist, which approach is more efficient in the long run?
3. How could we combine both methods next time? How could the AI group write a better prompt to force the tool to look for footnotes, data contradictions, and future-tense fluff?

Conclusion

In conclusion, this pilot study highlights a critical tension in how we interact with corporate sustainability disclosures and evolving technologies. While respondents in the pilot study have a strong, value-driven desire to align their financial portfolios with ethical and sustainable practices, the complexity of contemporary ESG reporting creates a significant barrier to independent understanding. To bridge this gap, users often use AI tools as processing assistants to summarise dense narratives and complex datasets. However, a profound trust gap remains as respondents exhibit low confidence in AI’s reliability for high-stakes investment choices and firmly reject blind reliance on automated outputs. Ultimately, respondents seek a hybrid model in which AI handles the exhaustive workload of data evaluation, leaving final ethical and financial judgments to human oversight. There is a gap between how we teach students to use AI and how they will need to use it, alongside the skills they are learning, once they graduate. As educators, we need to close that gap by demonstrating good practice in clear and effective ways. AI is an accelerator, but humans provide the critical oversight.

Special thanks go to The Open University for funding the pilot survey, and to my colleague, Devmali Perera (Lecturer in Finance, devmali.perera@open.ac.uk), for designing and creating the survey.

References

Cahan, S.F., De Villiers, C., Jeter, D.C., Naiker, V. and Van Staden, C.J. (2016) Are CSR disclosures value relevant? Cross-country evidence. European Accounting Review, 25(3), pp. 579–611. Available at: https://doi.org/10.1080/09638180.2015.1064009

De Villiers, C. and Van Staden, C.J. (2011) Where firms choose to disclose voluntary environmental information. Journal of Accounting and Public Policy, 30(6), pp. 504–525. Available at: https://doi.org/10.1016/j.jaccpubpol.2011.03.005

De Villiers, C., Rinaldi, L. and Unerman, J. (2014) Integrated Reporting: insights, gaps and an agenda for future research. Accounting, Auditing & Accountability Journal, 27(7), pp. 1042–1067. Available at: https://doi.org/10.1108/AAAJ-06-2014-1736

De Villiers, C., Dimes, R. and 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), pp. 96–118. Available at: https://doi.org/10.1108/SAMPJ-02-2023-0097

KPMG (2024) The move to mandatory reporting: Survey of Sustainability Reporting 2024. KPMG International. Available at: https://kpmg.com/xx/en/our-insights/esg/the-move-to-mandatory-reporting.html (Accessed: 5 November 2026).

Lewis, C. and Young, S. (2019) Fad or future? Automated analysis of financial text and its implications for corporate reporting. Accounting and Business Research, 49(5), pp. 587–615. Available at: https://doi.org/10.1080/00014788.2019.1611730

Lombardi, R. and Secundo, G. (2021) The digital transformation of corporate reporting – a systematic literature review and avenues for future research. Meditari Accountancy Research, 29(5), pp. 1179–1208. Available at: https://doi.org/10.1108/MEDAR-04-2020-0870

Misiuda, M. and Lachmann, M. (2022) Investors’ perceptions of sustainability reporting – a review of the experimental literature. Sustainability, 14(24), p. 16746. Available at: https://doi.org/10.3390/su142416746

Moodaley, W. and Telukdarie, A. (2023) Greenwashing, sustainability reporting, and artificial intelligence: a systematic literature review. Sustainability, 15(2), p. 1481. Available at: https://doi.org/10.3390/su15021481

Dr Katie Balaam

Katie is a Lecturer in Accounting and Financial Management at the Open University. She previously worked at the University of Birmingham and Queen Mary University of London. Katie also worked at the University of Buckingham for twelve years as a Senior Lecturer in Accounting. In 2018, Katie obtained her PhD in the field of corporate governance and capital structure from Birkbeck, University of London. Before moving to academia, Katie worked as a financial accountant in the automotive sector and is an FCCA-qualified accountant.

Part of the Pedagogy series
Join the Accounting Cafe community
© Accounting Cafe

How to cite this article: Balaam, K. (2026) ‘Want students to understand annual corporate sustainability reports? What are the challenges of using AI’, Accounting Cafe, [insert publication date]. Available at: PLACEHOLDER_ARTICLE_URL (Accessed: [insert date])