Behind the Scenes: Building an AI Tool for Sustainability Communications

Over the past few weeks, I’ve been deep in testing, refining, breaking, and rebuilding what will soon become the first live version of My Green Comms. Final tests are now running. Before I open it up more widely, I wanted to share a few honest reflections from behind the scenes.

Because building an AI tool specifically for sustainability communications has reinforced something I’ve believed for a long time: this isn’t really a writing problem. It’s a clarity and credibility problem. And those are much harder to fix with surface-level polish.

What I’ve Learned While Testing

When I started pressure-testing early versions, I deliberately compared outputs against generic AI tools. The results were interesting.

Generic AI writes fluently. It can produce confident, well-structured sustainability copy in seconds. But fluency isn’t the same as judgement. When I fed in draft sustainability claims, most models improved tone and flow, yet rarely questioned vague commitments or flagged potential risk. They didn’t ask whether the claim was defensible. They simply made it sound better.

In this space, that’s not enough.

The more drafts I reviewed, the clearer something became. The biggest issue wasn’t intentional greenwashing. It was vagueness. Language that wasn’t technically wrong, but wasn’t strong enough to build trust either.

Phrases like “we are committed,” “we aim to,” or “we strive to improve” appear everywhere. They feel safe. But they don’t create clarity. And clarity is what builds credibility, especially under increasing regulatory and stakeholder scrutiny.

Another pattern surprised me. The most valuable use cases weren’t requests like “write me a LinkedIn post.” They were far more strategic:

• Turn a 30-page ESG report into a coherent narrative.

• Map sustainability initiatives to business value.

• Identify where messaging could expose risk.

• Make complex technical content readable without oversimplifying it.

That’s where AI starts to become genuinely useful — not as a content generator, but as a thinking partner.

What’s Being Tested Now

Right now, I’m pressure-testing a set of focused tools designed around those real-world needs:

• A Greenwash Checker that flags weak or risky claims.

• A Message Mapper that helps clarify positioning before content is written.

• A Tone and Clarity Refiner that improves precision without over-promising.

• And structured prompts that connect sustainability language back to commercial relevance.

The guiding question throughout has been simple: how do we make AI helpful without making it careless?

Because in sustainability communications, carelessness carries consequences.

I’d Love Your Input

As this moves closer to launch, I’m keen to sense-check something with you.

If you could design an AI tool specifically for sustainability communications, what would you want it to do?

Would you want it to challenge weak claims?Translate technical ESG language into something more human?Stress-test messaging for regulatory risk?Build a strategic content plan from scratch?Something else entirely?

Early access will open soon. Those who join won’t just use the tool — they’ll help shape how it evolves.

If that sounds interesting, you can join the waitlist here.

More to come soon.