Read time: 6 minutes
Highlights
AI and genomics can help breeders find and develop resilient crop traits faster, but the varieties still need to prove themselves in the field.
- Tomatoes offer a clear signal: Heritable and Kagome are using AI to search for genetic traits linked to disease resistance.
- Editing is becoming more targeted: Biographica and Hudson River are combining AI-based gene discovery with crop-editing expertise.
- Strawberries are in the frame: Plantik and Hudson River are exploring heat-tolerant strawberries, building on work in tomato.
- Breeding remains essential: Finding a promising gene is only a starting point; growers need plants that perform reliably in real conditions.
- The possibilities are widening: Faster discovery could bring previously too-difficult crop traits back into scope.
Could artificial intelligence help breeders find a crop’s natural defences, then guide a precise change that helps it cope with disease or heat?
That’s one of the most compelling agriculture technology trends I’m watching. The goal isn’t simply to make a crop look new. It’s to help growers maintain dependable harvests as growing conditions shift and disease pressures change.
Recent work involving tomatoes and strawberries offers a valuable signal. It’s evidence that AI and genomics are being connected more deliberately to practical crop improvement.
From AI discovery to a stronger tomato
Heritable Agriculture and Japan’s Kagome are working together to find tomato genes associated with resistance to a major disease. Their collaboration brings together Kagome’s tomato breeding resources and research data with Heritable’s AI-based analysis. The partnership follows Kagome’s 2025 investment in Heritable, and the tomato work is the first phase of what both companies describe as a broader strategic program.
In plain English, AI helps search large collections of biological information for patterns that people might otherwise take a long time to spot. Researchers can look for relationships between a plant’s DNA, how its genes behave and the characteristics it displays. The result is a shortlist of possible genes for further study.
A disease-resistance gene is a piece of a plant’s genetic instruction set that can help it resist a pathogen, such as a fungus or virus. But genes don’t operate in isolation. A promising trait still needs to work alongside other important qualities, including yield, taste and the plant’s ability to grow well in different conditions.
One key detail is that the Heritable and Kagome project is focused on discovering resistance through genetic analysis and then using promising findings in breeding. It isn’t an announcement of a new, gene-edited tomato. The partners’ aim is to bring useful resistance into several Kagome varieties through breeding.
This is an important point. AI may help narrow the search, but plant breeders still need to test whether a promising trait translates into a productive, reliable crop.
What gene editing adds
A related development brings AI-guided discovery together with gene editing. UK-based Biographica and Dutch company Hudson River Biotechnology are working to connect the search for useful crop traits with the ability to make targeted changes to a plant’s genetic instructions.
Gene editing means making a specific change to DNA, not inserting new DNA. CRISPR-style gene editing tools are one way to do this. A useful plain-English comparison is a precise edit to a set of instructions, rather than rewriting the entire manual. Depending on the approach, the edit might switch a gene off or adjust how and when it works.
That second possibility could become especially valuable. Some genes that make a plant more vulnerable to disease may also support normal growth. Simply switching one off could create an unwanted trade-off. AI may help researchers predict which targets are worth exploring and how a change might affect other traits.
That doesn’t mean a computer can design a perfect crop on its own. It means models may help direct experiments towards promising options, while researchers and breeders assess what happens in the plant.
The partnership is initially focused on tomatoes, with the ambition of taking difficult crop traits from gene discovery towards edited plants. Their work makes another of the agriculture technology trends worth tracking especially clear. AI isn’t only being applied to farm machinery or decisions in the field. It’s also being used to help shape the crops themselves.
Why strawberries matter in a warming world
Tomatoes aren’t the only crop in this story. Hudson River is also working with Plantik on heat-tolerant strawberries, drawing on Plantik’s work identifying heat-tolerance traits in tomato and adapting its approach to strawberry.
The challenge is bigger than keeping a plant alive through a hot spell. Heat can disrupt flowering and fruit set, affecting how much fruit a plant produces and the quality of the harvest. A promising future variety would need to cope better with heat without compromising performance in more comfortable conditions.
Strawberries add a useful dimension to this emerging picture. Different crops can be difficult to edit and grow in the laboratory, and a method that works in tomato won’t automatically work in strawberry. Researchers still need to make progress crop by crop, then test whether the plants grow and produce fruit as intended.
The news is a significant signal, and trials and validation take time. But if the approach succeeds, it will offer breeders new ways to tackle traits that have been difficult to develop through traditional methods alone.
For me, this is where agriculture technology trends become strategically significant. The advances aren’t just about producing a faster research result. They could change which crop traits become feasible to pursue, and how soon businesses can explore them.
AI accelerates genomics, not certainty
In my TECHXponential™ framework, Genomics is one of five transformative technologies, alongside Alternative Proteins, Cellular Agriculture, The Microbiome and Synthetic Biology. All accelerated by four powerful forces, AI, Energy, Quantum Computing and Sensors.
Here, the connection between AI and Genomics is particularly visible. AI can help researchers sift through biological data, rank possible targets and decide which experiments to run next. Sensors and other measurement tools can then help collect information about plant performance. Over time, new results might help improve the next round of analysis.
That feedback loop could compress parts of crop development from decades down to years.
But we still need to consider biological complexity. A tomato that performs well in one setting might not perform the same way in another. And a change that helps with disease could have unintended effects on yield, quality or another part of the plant.
I’ve explored a related resilience question in Gene Edited Cacao: Can Genomics Stabilise Chocolate’s Future?. The crop is different, but the strategic question is similar. How could genomics help protect the agricultural foundations of products people rely on?
The broader backdrop is the one I set out in The New Front Line: How AI-Driven Crop Breeding is Reshaping National Sovereignty. That piece made the sovereignty case for AI-driven breeding. This is what that argument looks like at crop level, one company, one crop, one resistance trait at a time.
For more detail on the tomato collaboration, read AgNavigator’s report on Heritable and Kagome.
Future possibilities
The most exciting possibilities sit beyond any single crop or company. If AI-guided discovery keeps proving itself, the range of traits breeders can realistically pursue could widen considerably. Disease resistance that once took decades of backcrossing will become a targeted, testable objective. Heat tolerance, currently one of the hardest traits to breed for, could follow the same path the strawberry work is now exploring.
That would reshape what’s worth attempting. Crops that breeders have quietly written off as too difficult, too slow or too genetically complex can come back into scope. Ingredient portfolios can diversify as previously fragile crops become dependable again. And the gap between identifying a useful gene and growing a commercial variety will shrink from decades towards years, which changes who can participate in crop improvement at all.
The second-order effects are just as interesting. If crop improvement accelerates, the strategic value of a breeding program shifts from accumulating internal knowledge to owning the best data and the fastest feedback loop. Seed companies, food manufacturers and even software firms will then all compete for that position. And as edited varieties move closer to market, the questions that follow won’t only be technical ones about performance, but commercial ones about access, ownership and trust.
The direction of travel is clear, and the companies paying attention now will be the ones best placed to move when these possibilities start arriving.
The crop may become a technology platform
The most exciting possibility isn’t a perfect tomato or an invulnerable strawberry. It’s a future where breeders can explore more targeted solutions to specific crop challenges, and where AI helps them decide which routes are most worth testing.
That will create new opportunities for growers, food manufacturers and consumers. It can help protect familiar ingredients from mounting pressure. It also raises important questions about oversight, access, ownership and trust, all of which need to be part of the conversation.
Agriculture technology trends like these invite us to look further ahead than the next growing season. If AI can help identify useful traits faster, and genomics can help breeders work with them more precisely, what crop challenges might become solvable that aren’t today?
To continue the conversation email me at tony@futuristforfood.com
