2 min read
Greenwashing Watchdog: Checking Green Claims Against EU and UK Rules with AI
EU law now bans generic green claims like eco-friendly unless proven. I built Greenwashing Watchdog with AI agents to check a brand's claims in minutes.
Is it a legit green claim or greenwashing? Most brands only find out when the fine arrives.
Since 27 September 2026, EU law bans generic green claims like "eco‑friendly" and "green" unless a brand can prove them, with fines of at least 4% of annual turnover in cross‑border cases. UK regulators can already fine up to 10% of worldwide turnover. This is not theoretical. Eni has paid €5 million in fines for the words "Green Diesel".
So I built a tool for it.
Generic green claims are now a legal risk
The words themselves are not new. "Green", "eco-friendly", "sustainable" have been sitting on packaging, product pages, ads and social posts for years. Nobody thought twice about them. What changed is that the same word now needs proof behind it, and the bill for getting it wrong is calculated on turnover, not on the size of the mistake.
The hard part is not fixing a claim. The hard part is knowing which claims you have, where they are, and which ones would not survive a regulator reading them. In most companies nobody has that list. The claims were written by different people, in different years, for different campaigns. That is exactly why most brands only find out a claim is misleading when the fine arrives.
What Greenwashing Watchdog does
Name any brand and Greenwashing Watchdog:
- checks its green claims against UK and EU rules, in minutes
- shows where each claim was found, so you can see the source and not just a verdict
- flags the misleading ones
- suggests a version of each claim that holds up
The goal is simple: fix it before a regulator does.
Built with AI agents at Claude Founder House Stockholm
I built Greenwashing Watchdog in Stockholm, where I was for Anthropic's Claude event, at Claude Founder House. It runs on Claude Managed Agents by Anthropic, and I built it together with Grok Bot.
This is the kind of work where AI agents are genuinely useful today. Reading a lot of public text, comparing it against rules, keeping track of where every statement came from, and rewriting it in a better way. It is text in, text out, and a lot of it. A person would need days to do it for one brand. An agent does it while you wait.
That does not mean you skip judgment. It means the boring part, finding and listing the claims, is no longer the bottleneck. You spend your time deciding what to change, not hunting for what to check.
Try it for free
Greenwashing Watchdog is free to try: Greenwashing Watchdog. Pick a brand, yours or someone else's, and see what comes back.
If you were in Stockholm that week and we did not get that coffee, my DMs are still open.
