Greenwashing is not always intentional. Sometimes it begins with a confident sentence in a report, a vague sustainability claim in a presentation, or a project update that sounds better than the available evidence.
A team may say a project is “sustainable,” “inclusive,” “low-impact,” or “responsible”, but can they prove it?
This is where project managers have an important role. They are close to the work, close to the evidence, and close to the decisions. And increasingly, they can use AI to help check whether ESG claims are supported by facts.
The question is not whether AI can make a project sustainable but the question is can AI help project teams notice weak claims before they become greenwashing?
Greenwashing Often Starts with a Gap
Greenwashing happens when there is a gap between what is claimed and what can be demonstrated.
In project work, this gap can appear in many ways:
- a report mentions environmental benefits without showing data;
- a supplier is described as responsible without proper documentation;
- stakeholder engagement is claimed, but meeting records are incomplete;
- a project says it supports inclusion, but accessibility was never reviewed;
- carbon reductions are presented without explaining assumptions;
- risks are described as “managed,” but there is no evidence of follow-up action.
These gaps may not be dramatic at first. They may look like small shortcuts. But over time, they can damage trust.
For project teams, the danger is clear: ESG communication can move faster than ESG evidence.
That is why avoiding greenwashing is not only a branding issue. It is a project management issue. It depends on how teams collect evidence, review claims, record decisions, and escalate uncertainty.
How AI Can Support Better ESG Claims
AI can help project teams by acting as a critical reader.
Before a sustainability statement is published, reported, or shared with stakeholders, AI can help ask useful questions:
- Is this claim specific or too vague?
- Is there evidence to support it?
- Are the numbers consistent across documents?
- Are assumptions clearly explained?
- Does the language exaggerate the result?
- Are there missing risks or limitations?
- Is the claim aligned with what the project actually delivered?
For example, a project report might say:
“The project significantly reduced environmental impact.”
An AI-supported review could flag this as too broad and ask:
- What impact was reduced?
- Compared to what baseline?
- By how much?
- Over what period?
- Where is the evidence?
- Who verified the data?
This does not mean AI decides whether the claim is acceptable. It means AI helps the team slow down and check the statement before it becomes part of official communication.
In this way, AI can support more honest ESG language.
Instead of saying “significantly reduced impact,” the team might write:
“The project reduced estimated electricity consumption by 12% compared with the previous system, based on internal monitoring data from the first three months of operation.”
That sentence is less dramatic. But it is much more credible.
AI as an Evidence Checker, Not a Truth Machine
It is important to be realistic. AI is not a truth machine.
AI can summarise documents, compare statements, detect inconsistencies, and suggest missing evidence. But it can also misunderstand context, overlook important details, or produce confident but incorrect outputs.
That is why AI should be used as a support tool, not as the final authority.
A useful approach is to give AI a clear role in the review process. For example:
Step 1: Collect the claim
What exactly are we saying about the ESG impact of this project?
Step 2: Link the evidence
Which documents, indicators, calculations, or records support the claim?
Step 3: Run an AI-assisted review
Ask AI to check for vague wording, unsupported statements, inconsistent numbers, and missing assumptions.
Step 4: Human review
The project manager, ESG expert, or responsible decision-maker checks the AI feedback and decides what to change.
Step 5: Record the decision
The final wording, evidence source, and approval should be documented.
This creates a practical balance. AI helps with speed and attention to detail. Humans remain responsible for judgment and accountability.
What Project Managers Should Watch For
Project managers do not need to become legal experts or data scientists to reduce greenwashing risk. But they do need to build a healthy habit of asking evidence-based questions.
Here are five useful checks:
1. Check the wording
Words like “green,” “sustainable,” “responsible,” “ethical,” or “impactful” can be useful, but only if they are explained. If a claim sounds impressive but unclear, it probably needs more detail.
2. Check the evidence
Every ESG claim should connect to something concrete: data, documents, decisions, meeting notes, supplier records, audits, or project outputs.
3. Check the boundaries
What is included in the claim, and what is not? A project may reduce paper use but increase energy use. A balanced statement should not hide relevant trade-offs.
4. Check the consistency
Numbers and claims should match across presentations, reports, dashboards, and public communication. AI can be especially helpful in finding inconsistencies across many documents.
5. Check the accountability
Who approved the claim? Who owns the data? Who can explain the assumptions? If nobody owns the statement, the statement is a risk.
These checks can become part of normal project routines. They do not need to be complicated. They simply make ESG communication more disciplined.
A Simple Example: Before and After AI Review
Imagine a project team preparing a final report.
The first draft says:
“Our project created a greener and more socially responsible solution for the community.”
It sounds positive, but it is too broad. An AI-assisted review might flag several issues:
- “Greener” is not defined.
- “Socially responsible” is not explained.
- The community impact is not supported by evidence.
- No indicators or examples are provided.
After review, the team rewrites it:
“The project introduced a digital service that reduced printed forms by approximately 8,000 pages during the pilot phase. It also included two user-testing sessions with local community representatives, which led to three accessibility improvements before launch.”
This version is not perfect, but it is much stronger. It is specific, evidence-based, and easier to verify.
That is the practical value of AI in avoiding greenwashing. It helps teams move from attractive language to accountable language.
Conclusion
AI cannot make ESG claims true. Only real project actions, reliable data, and responsible decisions can do that.
But AI can help project teams avoid weak, vague, or unsupported statements. It can act as a second pair of eyes, a consistency checker, and a prompt for better questions.
For project managers, this is a valuable opportunity. Greenwashing prevention does not have to happen only at the end of a project, when the report is almost finished. It can be built into daily project work: during planning, evidence collection, risk reviews, stakeholder communication, and closure.
The goal is not to make ESG communication less ambitious. The goal is to make it more trustworthy.
So, before your next project report says something is “sustainable,” ask:
Can we show the evidence, and would we be comfortable explaining it to stakeholders?
If the answer is yes, the claim is stronger.
If the answer is no, AI can help you find the gap before someone else does.
Marina Krstić is a Project Manager at Sparky* and co-founder with extensive experience in European projects and the management of complex, AI-driven initiatives. Her work focuses on bridging technology, sustainability, and business value, with a particular emphasis on delivering large-scale AI solutions in the finance and media industries.
#ESG #ArtificialIntelligence #SMEs #Sustainability #ProjectManagement #AIAgents #ESG4PMChange #DigitalTransformation