When a company uses a digital tool to evaluate its sustainability stance, a key question arises: if the tool generates a result, can the process behind it be explained? For smaller enterprises, this concern is quite concrete. An unexplained figure is hard to defend when a client or lender asks about it.
Where the data challenge begins
Part of the challenge lies in the nature of the information itself. The OECD notes that sustainability data are often expensive to produce, hard to verify, and scattered across incompatible reporting frameworks. AI and digital tools can be responsibly utilised only under appropriate governance conditions (OECD, 2026b). Therefore, transparency does not stem solely from technology; it depends on the quality of the underlying information and how the process is organized.
Practical signs of a transparent assessment
A transparent AI-supported assessment should provide a clear evidence trail. Users should be able to identify the data that contributed to the result, the reporting period and organisational boundary, any assumptions or estimates, and the limitations of the output. The result should be open to review rather than presented as an indisputable conclusion. These characteristics align with TRUST-ESG's focus on ESG information that is accessible, traceable, and explainable.
Questions an SME should be able to ask
One practical method to evaluate the transparency of an AI-supported outcome is to assess whether the business can answer a few key questions about it:
- What data contributed to this result?
- Is any relevant information missing?
- Were any assumptions used to produce it?
- Can the result be traced back to its source?
- Can an individual review or challenge the outcome?
- Is the result intended to support decision-making rather than serve as an indisputable conclusion?
If these questions can be answered, the basis of the result is apparent, which is what transparency entails in practice. If they cannot, the result may still be produced, but it becomes harder to trust and defend.
Transparency is not the same as accuracy
One distinction deserves emphasis: transparency and accuracy are related but not identical. A transparent process makes the basis of a result visible: the data, the steps, the assumptions, allowing it to be examined. It does not, by itself, ensure the result's correctness. A figure can be fully traceable yet still incorrect if the underlying data was flawed. Transparency lets you scrutinise how a result was reached and, consequently, verify it. Transparency should not be confused with complexity: a simple, well-documented assessment with traceable reasoning is more transparent than a sophisticated one with concealed workings.
What does this look like for an SME
Consider a small company using a digital tool to estimate its annual carbon footprint. A transparent process would allow the business to see which figures contributed to the estimate: energy bills, fuel records, perhaps travel data and which were estimated rather than measured. It would present any assumptions the tool applied, such as a standard emission factor, and highlight where information was incomplete. If the estimate later appears unusually high, the business can trace it back, identify the cause, and correct it. If the same estimate arrived as a single number with no visible basis, the company would have no choice but to accept or reject it without insight. The difference lies not in the tool's sophistication but in whether its reasoning is visible.
This is the standard TRUST-ESG seeks to uphold: technology should not merely deliver an answer. It should help an SME understand the evidence behind that answer, recognize its limitations, and communicate it with confidence. A simple transparency check involves selecting one AI-supported result and confirming that you can demonstrate its source data, assumptions, and limitations.
References
OECD. (2026b). Leveraging AI and digital tools for SME sustainable finance (OECD SME and Entrepreneurship Papers No. 80).