AI and Sustainability Trends 2026: 10 Takeaways from Our Barometer

Generative AI reached sustainability teams faster than most tools before it. In 2025, the first edition of our study already showed daily use, but little structure around it. A year later, use is widespread. The rest has moved less: skills, budgets and internal rules.

The AI x Sustainability Barometer 2026 surveyed more than 100 sustainability, QHSE and compliance professionals between July and September 2026. This second edition lets you compare the results with 2025. Here are the ten trends that stand out.

1. AI's environmental impact becomes the top barrier

This is the sharpest reversal of the year. In 2025, the environmental impact of AI was the least-cited barrier. In 2026, it comes first: 75% of sustainability professionals who use AI little or not at all cite it, ahead of data privacy (60%).

Sustainability teams know the paradox well: they measure their company's footprint, and the tool that saves them time adds to it. The barometer breaks this barrier down by sector, and shows what respondents expect from AI providers.

2. Adoption is moving faster than skills

93% of sustainability professionals use an LLM, up from 80% in 2025. Yet more of them than a year ago rate themselves as beginners or basic users: 77%, up from 69%.

That gap has a cost, because time savings grow with skill. Intermediate users save far more time each week than beginners. Training up, for example with a short guide to AI prompts, pays off quickly.

3. Nine in ten companies have no AI budget for sustainability

93% of companies have no approved budget for AI tools dedicated to sustainability and compliance. Fewer than one in four has a formal internal AI policy.

In other words, AI is already in the team, but the company hasn't organized it yet. Employees use consumer tools, each in their own way. The barometer also looks at EU AI Act compliance, and the picture is far from reassuring.

4. The finance & tech sector pulls ahead of manufacturing

Not every sector moves at the same pace. In finance and tech (grouped into a single category), 93% of sustainability professionals use AI daily or weekly, compared with 56% to 67% in other sectors. They also test AI agents far more, and are twice as likely as the manufacturing sector to expect a major impact within three years (71% vs. 37%).

The manufacturing sector trails on almost every one of these indicators. The barometer details the gaps for each sector.

5. Still optimistic, but less than in 2025

Sustainability professionals still trust AI: 57% see it as an opportunity, while 25% see it as a risk. But enthusiasm has cooled. The share expecting a major or transformative impact on their sector within three years fell from 60% to 53%.

Part of the promise has been tested, and it didn't always deliver. The first-hand accounts in the barometer describe what disappointed respondents most.

6. When senior management owns AI, use of AI agents takes off

Who owns AI matters a lot. When it's senior management or the executive committee, 57% of companies use or test AI agents. When it's QHSE alone, only 23% do.

Across all respondents, use of AI agents rose from 5% in 2025 to 37% in 2026. Senior management involvement goes a long way toward explaining that scale-up.

7. Large companies save more time, small ones test more agents

At companies with more than 2,000 employees, 92% of sustainability professionals save at least two hours a week thanks to AI. Below 500 employees, it's one in two.

Yet companies with fewer than 101 employees adopt AI agents the most (58%). Size doesn't always push in the same direction. The barometer details the results by company size.

8. Sustainability leaders are more convinced than their teams

Role shapes perception too. Two in three sustainability managers and directors see AI as an opportunity (67%). Among sustainability officers, it's fewer than one in two (46%).

Those who lead the program see what AI can change for the whole team. Those who do the day-to-day work see first what needs checking and fixing. Driving adoption means convincing both groups, and not with the same arguments.

9. Nine in ten check what AI produces

90% of sustainability professionals always or often check AI output before using it. Sustainability managers and directors check it systematically 82% of the time, officers 68% of the time.

That makes sense when a mistake can end up in a public report. 58% also believe AI increases the risk of greenwashing in sustainability reporting, up from 53% in 2025.

10. AI is a communication tool first, not yet a compliance tool

Sustainability professionals mostly use AI to write: emails, visuals and translations (71%), report sections (62%). Yet data collection is the most time-consuming task for half of them, and only 13% hand it over to AI.

So AI writes the non-financial report, but still does little to prepare the data behind it. The biggest time savings, though, are in data collection, supplier follow-ups and EcoVadis questionnaires. That's the job of AI solutions for sustainability compliance, built for those tasks rather than for writing alone.

What these trends say about 2027

Whether sustainability teams use AI is no longer the question. The question for 2027 is governance: who owns it, with what budget, for which tasks and within which environmental limits. Companies that answer early are also the ones getting the most out of it.

The full barometer goes further than these ten trends. It includes results by sector, company size and role, where companies stand on the EU AI Act, and first-hand accounts from sustainability professionals on how they use AI and where they draw the line.

AI and Sustainability Trends 2026: Key Takeaways