A Zero Waste Llama survival guide
If you’ve spent any time around sustainability marketing, you already know the pattern: lots of buzzwords, vague claims, and very little substance. The conversation around green AI greenwashing is becoming increasingly relevant in these discussions.
AI is no different — it just wears a shinier jumper.
Use this checklist whenever a company claims its AI is “sustainable”, “eco‑friendly”, or “part of the climate solution”.
🦙 1. Ask: Green compared to what?
If a product says it “reduces emissions” but doesn’t say compared to what baseline, that’s a red flag.
- Reduced compared to doing nothing?
- Compared to an older system?
- Compared to a hypothetical worst‑case scenario?
Zero Waste Llama rule:
If there’s no baseline, there’s probably no reduction — just creative framing.
🦙 2. Watch for efficiency hype without limits
Phrases to be suspicious of:
- “Optimised”
- “Smarter”
- “More efficient”
- “Streamlined”
Efficiency can be useful — but only if total energy and resource use goes down, not just cost per task.
If efficiency leads to more usage, more scale, and more growth, it’s not sustainability. It’s rebound effect.
🦙 3. Look for what’s missing: energy, water, hardware
Most “green AI” claims talk about outputs — rarely about inputs.
Ask:
- How much energy does this system use (training + daily use)?
- Where does that energy come from, hour by hour?
- How much water is used for cooling?
- What hardware does it require, and how often is it replaced?
If these aren’t mentioned at all, they’re being quietly ignored.
🦙 4. Be wary of “runs on renewable energy” claims
This one sounds great — until you squint a bit.
Common tricks include:
- Using annual averages instead of real‑time data
- Relying on carbon offsets instead of actual reductions
- Claiming renewables at company level, not project level
Zero waste thinking reminder:
Offsets don’t remove impact — they just move responsibility elsewhere.
🦙 5. “AI for sustainability” is not the same as “sustainable AI”
This is a big one.
- AI used to analyse environmental data ≠ low‑impact AI
- AI that supports climate work can still have a high footprint
If a company is solving one sustainability problem while creating another (energy use, water stress, e‑waste), that trade‑off needs to be openly acknowledged — not buried.
🦙 6. Vague language = deliberate vagueness
Watch for phrases like:
- “Helping the planet”
- “Driving positive change”
- “Supporting net zero goals”
- “Enabling sustainability journeys”
Then check:
- Are there numbers?
- Are there limits?
- Are there independent assessments?
If everything is aspirational and nothing is measurable, you’re looking at tech vibes, not climate action.
🦙 7. No lifecycle? No sustainability claim.
A genuinely sustainability‑aware AI product should address:
- Extraction (minerals, materials)
- Manufacturing
- Energy and water use
- Hardware lifespan
- End‑of‑life disposal
If the conversation starts and ends at “look what the software can do”, it’s only telling half the environmental story — at best.
🦙 8. Ask the uncomfortable question: Do we actually need this?
True zero‑waste thinking always includes the option not to build.
Ask:
- Is this replacing something impactful, or just adding another layer?
- Could this be done with simpler tools?
- Is AI being used because it’s necessary — or because it attracts funding and attention?
Not every problem needs AI.
Sometimes the greenest option is… less.
🦙 9. If growth is the goal, sustainability is not
If a company’s business model depends on:
- Rapid scaling
- Constant expansion
- Increasing data use
- Always‑on deployment
Then any environmental benefit needs to be treated with caution.
Infinite growth + finite planet is still the problem — AI or not.
The Zero Waste Llama takeaway
Green AI isn’t about prettier dashboards or smarter algorithms.
It’s about honest accounting, real limits, and knowing when not to optimise.
If a sustainability claim:
- Acknowledges trade‑offs
- Shows restraint
- Shares uncomfortable data
- Accepts limits
— it might be worth listening to.
If it sounds too clean, too easy, or too good to be true?
The llama recommends keeping your sceptical eyebrows raised. 🦙💚
This post was researched & written to provide a checklist for professionals to refer to before using AI. Is AI Really Green?
Key Takeaways
- Beware of vague claims in sustainability marketing; many companies may engage in green AI greenwashing.
- Ask clear questions about AI claims, such as what baseline they compare to and what resources they use.
- Watch for efficiency claims that do not lead to real reductions in energy or resource use; these can be misleading.
- Understand that ‘AI for sustainability’ does not equal ‘sustainable AI’; trade-offs need transparency.
- Evaluate claims critically; if it sounds too good to be true, stay sceptical and demand concrete data.
