Artificial intelligence offers SMEs new opportunities to make sustainability reporting more efficient. It can be used to collect and structure sustainability data, interpret reporting requirements and prepare report texts considerably more efficiently than before. However, AI cannot replace reliable source data, appropriate calculation methods or a company's responsibility for the accuracy of the information it reports. How can SMEs use AI in sustainability reporting effectively and safely?
AI Lowers the Barrier to Sustainability Reporting
Sustainability reporting is often perceived by SMEs as a laborious process requiring specialized expertise, complex calculations and extensive data collection. However, the biggest challenge is often not writing the report itself, but identifying, collecting and processing the necessary information.
A company's sustainability data is typically scattered across multiple sources. Energy consumption data can be found on utility bills, employee information in HR systems, procurement data in financial management systems, and sustainability-related policies in various internal company documents. Bringing all this information together into a coherent whole can be time-consuming.
AI can make this work easier. It is particularly well suited to processing existing information, analyzing documents, classifying data and preparing sustainability reports. At its best, AI frees up company resources for actual sustainability work – improving business operations.
The use of AI in sustainability reporting has also gained attention in Finnish business development initiatives. In the VAHVA project carried out by JAMK University of Applied Sciences between 2024 and 2026, the use of AI to support sustainability reporting was one of six operational models developed. The project's experiences demonstrate that SMEs primarily need practical tools and expertise to improve their sustainability reporting.
The New EU Voluntary Sustainability Reporting Standard Also Provides a Clear Framework for the Use of AI
The new EU voluntary sustainability reporting standard, which entered into force in September 2026 and was established by Commission Delegated Regulation (EU) 2026/1560, provides SMEs with a common framework for presenting sustainability information. The standard builds on the previous VSME framework and retains its two-part structure: the Basic Module and the Comprehensive Module.
The standard aims to make it easier for companies to respond to the sustainability information needs of customers, financiers and other stakeholders. At the same time, it helps companies develop their own sustainability practices and identify sustainability issues relevant to their business operations.
AI can be particularly useful when applying the standard. For example, a company can use AI to identify the reporting requirements set out in the standard, compare them with information already available within the company and compile a list of missing data.
In practice, AI can help create a reporting table showing each disclosure requirement, the information available within the company, the relevant data source and any need for further investigation. This makes the reporting process more systematic and helps companies focus their resources on the most relevant issues.
However, it is important to provide AI with the current version of the standard as source material and verify the accuracy of its interpretations. Not all data points in the standard apply equally to every company. Some disclosures are conditional, others are voluntary, and certain simplifications have been introduced for companies with no more than ten employees.
AI can help identify reporting requirements, but the company must verify which information is actually required under its chosen reporting scope.
Where Can AI Provide the Greatest Benefits?
The use of AI in sustainability reporting can be divided into three practical stages: collecting data, structuring information and preparing the report.
1. Collecting Sustainability Data
AI can process a company's existing documents and help identify information needed for sustainability reporting. For example, it can extract energy consumption data from utility bills, employee numbers and occupational safety information from HR reports, and descriptions of sustainability governance from company policies.
This does not mean that AI automatically collects data from all of a company's systems. Data availability depends on the applications used, their integrations and access permissions. In addition, information extracted by AI must be verified against the original sources.
When implemented effectively, AI-assisted data collection can reduce manual work and make preparing a company's first sustainability report considerably easier.
2. Structuring Information According to the Reporting Standard
Once the necessary data has been collected, AI can help organize it under the appropriate reporting requirements. The Basic Module of the EU voluntary standard includes, for example, disclosures relating to energy and greenhouse gas emissions, water use, resource use, employees and business conduct.
AI can compare a company's data with the structure of the standard and identify potential gaps or inconsistencies. For example, it may flag cases where energy consumption has only been reported as costs in euros, even though the reporting requirements call for consumption figures in megawatt-hours.
AI can also help develop data collection templates for future reporting periods. This allows companies to establish a repeatable reporting process rather than collecting information from scratch every year without a consistent framework.
3. Preparing the Report Text
Writing the report is one of the most natural applications of AI. Once a company's sustainability data, objectives and actions have been defined, AI can help turn them into clear and understandable report text.
For example, AI can transform a technical description of energy consumption into reader-friendly language, standardize terminology throughout the report, summarize lengthy sections or produce alternative presentations for different target audiences.
This is particularly useful for small businesses, where the person responsible for sustainability reporting often handles the task alongside their other responsibilities.
However, well-written report text does not necessarily mean that the information it contains is reliable. This is where the most important limitation of AI becomes apparent.
AI-Generated Text Is Not the Same as Verified Sustainability Information
Generative AI produces responses based on the data it processes and the linguistic patterns it has learned. It can write a convincing description of a company's sustainability performance even when the available source data is incomplete. This is precisely what makes the use of AI in sustainability reporting potentially risky.
For example, if AI is asked to describe a company's climate impacts without actual energy consumption and emissions data, it may produce a generic text about the company's efforts to reduce emissions. The text may sound credible even if the company has no defined emissions reduction targets or information about its actual emissions.
In sustainability reporting, it is therefore essential to distinguish between three elements:
- Source data: information collected from the company's operations, such as energy consumption, fuel use, employee numbers and occupational accidents.: yrityksen toiminnasta kerätyt tiedot, kuten energiankulutus, polttoaineiden käyttö, henkilöstömäärä ja työtapaturmat.
- Calculation results: results derived from source data using documented methods, such as greenhouse gas emissions.
- Report text: a written description of the company's sustainability performance based on source data and calculation results.
AI can assist in all three stages, but its output does not in itself verify the accuracy of source data or the validity of calculations. The reliability of a sustainability report depends on verifiable information and transparent methods, not on how convincingly the report is written.
Can AI Calculate a Company's Carbon Footprint?
Carbon footprint calculation is a good example of a task where the opportunities and limitations of AI must be understood. AI can help identify the necessary source data, organize consumption data, create calculation spreadsheets and check the logic of calculations. It can also explain calculation methods and help document the calculation process.
However, calculating a carbon footprint requires accurate source data, appropriate emission factors and a consistent calculation methodology. For example, calculations based on the GHG Protocol must account for the classification of emission sources as well as the organizational and operational boundaries of the assessment.
AI cannot reliably determine a company's greenhouse gas emissions based solely on its industry, turnover or number of employees. Estimates based on such information may be used as indicative starting points in certain situations, but they must not be presented as measured or accurately calculated emissions.
Emission factors suggested by AI must also be verified. The suitability of an emission factor depends on factors such as the method of energy production, geographical location, reporting year and calculation methodology used.
AI can be used as a technical tool to support calculations, but the accuracy of the results must be verified separately. A company's carbon footprint does not become reliable simply because the calculation was produced by AI.
Data Security and Confidentiality Must Be Taken into Account
Sustainability reporting also involves information that a company may not wish to disclose to external parties. This may include supply chain information, energy costs, employee data and internal business development plans.
Before uploading materials to an AI service, it is important to establish how the service processes data, where the data is stored and whether it may be used to train the service provider's models. The processing of personal data must also comply with data protection legislation.
Companies should establish clear policies for the use of AI. Not all information needs to be, or should be, uploaded to an AI service in its original form. Anonymized, aggregated or otherwise limited data is often sufficient for preparing sustainability reports. Secure data handling is an essential part of responsible AI use.
How SMEs Can Get Started in Practice
Using AI in sustainability reporting does not require major software investments or complex IT projects. Companies can get started with small, manageable steps.
First, the company defines its reporting objectives and selects an appropriate reporting scope. The next step is to identify the necessary information and determine where it can be obtained. At this stage, AI can be used to help prepare a data collection plan and reporting framework.
Next, the company collects its actual sustainability data and performs the necessary calculations. For each reported indicator, the data source, reporting period and, where applicable, calculation methodology are documented. Missing information is identified as such rather than allowing AI to fill the gaps with assumptions.
Only after these steps is AI used to prepare the actual report text. Finally, the company verifies that all statements, figures and targets presented in the report are consistent with documented source data and decisions made. This allows AI to support the reporting process without the company losing control over the content of its report.
The Greatest Benefit of AI Comes from Better Data Management
The value of AI in sustainability reporting should not be assessed solely by how quickly it can produce a report. A much greater benefit may come from helping companies collect, structure and utilize their sustainability data more systematically than before.
When energy consumption, material flows, employee-related indicators and other sustainability metrics are monitored regularly, the information also becomes a valuable tool for business management. It can help identify cost savings, operational risks and opportunities for improvement, as well as track progress towards established targets.
AI can help transform scattered information into a coherent and useful whole. However, it cannot determine on behalf of a company which sustainability issues are most relevant to its business or which targets the company should commit to. These decisions require an understanding of the company's own operations, business environment and stakeholder expectations.
AI does not make a company sustainable or a sustainability report reliable. When used appropriately, however, it can make sustainability management more efficient, systematic and easier to manage. For SMEs, this is a far more valuable objective than simply producing a sustainability report quickly.
Sources and Further Reading
- Euroopan komissio (2026). Komission delegoitu asetus (EU) 2026/1560. Vapaaehtoista vastuullisuusraportointia koskeva standardi. https://eur-lex.europa.eu/legal-content/FI/TXT/?uri=CELEX:32026R1560
- Jyväskylän ammattikorkeakoulu (2026). VAHVA-hanke vahvisti keskisuomalaisten pk-yritysten valmiuksia vastuullisuusraportointiin. https://www.jamk.fi/fi/uutiset/2026/vahva-hanke-vahvisti-keskisuomalaisten-pk-yritysten-valmiuksia-vastuullisuusraportointiin
- Jyväskylän ammattikorkeakoulu. VAHVA – Vastuullisuusraportoinnilla vahvaan ja kestävään kasvuun Keski-Suomessa. https://www.jamk.fi/fi/tutkimus-ja-kehitys/tki-projektit/vahva-vastuullisuusraportoinnilla-vahvaan-ja-kestavaan-kasvuun-keski-suomessa
- GHG Protocol. Corporate Accounting and Reporting Standard. https://ghgprotocol.org/corporate-standard
Käkikuu Oy helps SMEs develop sustainability practices that support their business operations, identify material sustainability issues and produce reliable sustainability information. Explore our services at www.kakikuu.fi.
EN 
FI