World Agri-Tech London 2026, Day 2: AI in Agriculture, Excitement and a Healthy Alarm
Day 2 of the World Agri-Tech Innovation Summit was about AI in agriculture: 400 crop-robotics startups, quick wins in the packhouse, the data-ownership question, and why the risks of AI call for awareness, not panic.
World Agri-Tech Party – Day 2.
We spoke a lot about AI in Agriculture, and I have such dual feelings about it right now. Two days in London only amplified them.
The first feeling: excitement
The first feeling is obvious: excitement. AI is already bringing a lot of positive change to farming. This was confirmed at the expert panel on autonomous and precision farming and at a 40-minute workshop on Agentic AI in Agriculture.
The headliner of both was Dr Elliott Grant, who knows the subject not from the news, but from the depth & inside.
A few insightful things:
1. A very dynamic market
Around 400 ag automation startups are on the latest crop robotics landscape.
2. The drivers are real
I hear them on my own Zoom calls with farmers, especially in the Global South: labour shortages, rising input costs, pressure on pesticides.
3. Quick AI wins are simple
A robot that just moves fruit around a packhouse frees the best pickers to keep picking.
4. Who owns the data?
Farmers run mixed fleets of different brands, so interoperability and "who owns the data?" come up every time. From my own experience: in most cases your tech is co-existing with the tech zoo already deployed on the farm.
5. Britain is still a bit British about it
And on AI in UK agriculture: Britain is still a bit British about it (from the stage, not from me :)
The second feeling: a healthy alarm
So, the excited part is finished. Now the second, healthy feeling. Not fear, but inner alarm about the risks of AI.
You know that feeling when you meet a person with amazing depth of knowledge and unbelievable speed of work... but their emotional intelligence is still in its infancy? You don't know how to behave with that person...
Three scenarios if something goes wrong with AI (Geoffrey Hinton, the godfather of AI, said this, I'm just repeating):
- Bad guys use AI on purpose & create new global biological risks.
- Smart, good guys use it inaccurately and create something not in the plan (negligence; check the Hugging Face incident, brilliantly explained by Elliott 🤗).
- AI becomes smarter than us and decides it doesn't need humans anymore.
So what to do? As usual, don't panic :)
- Be aware of the risks, but focus on your own universe, your best farm or agrifood production.
- Be aware of the risks, but keep moving forward with AI step by step.
- More people want AI for good: productivity, less routine, easier lives for growers.
- There is huge progress in good AI with vibe coding, 3D printing and GenAI, even in business pipelines.
- In the last month alone I've seen at least a dozen GitHub repositories where someone builds physical AI just for fun, but with obvious commercial potential. So, let's not be scared, but aware.
- You are not alone, we are all in this unclear time together.
The only absolutely clear thing
The era of permissionless AI innovation in agriculture is happening right here, right now 🚀
And I felt it at World Agri-Tech 💚
Day 1 of the summit, on what is actually slowing agritech down, is here.