Episode 67. Fixing Food’s Broken Economics with Sophia Fannon-Howell from Aterra.ai
AI Will Matter in Food When It Shifts Power, Not Just Productivity
Sophia Fannon-Howell on why farmers need better data, stronger market access, and a bigger share of the value they create
For all the excitement around AI in agriculture, the real opportunity may have less to do with growing more food and far more to do with fixing who holds the power. Sophia Fannon-Howell argues that the UK food system is structurally imbalanced: farmers carry the risk, produce the value, and yet retain only a fraction of the reward, while larger players control the data, the market visibility, and the negotiating leverage.
That imbalance is what led her to found Aterra, a venture focused on using technology to support a fairer, more regenerative food system. Her view is that AI should not simply be layered onto farming as another productivity tool. Used properly, it can become a way to give farmers better intelligence, stronger bargaining power, and more direct access to markets.
From energy and commodities to food systems
Sophia’s route into this work began in the energy sector. With a doctorate in geology, she entered the industry expecting a more traditional technical path, but low oil prices pushed her instead into data-focused roles. That early shift shaped the direction of her career.
She went on to build deep experience in data management, digital transformation, and analytics across energy and, later, commodity trading. Working in those sectors gave her a close view of how large organisations use data and AI to optimise supply chains and capture value. Over time, that perspective led her to a larger question: if these tools are so powerful in corporate settings, why are they not being used to address deeper structural problems elsewhere in the economy?
Food and farming became the answer. In her view, they sit alongside energy as foundational systems, and once she began looking at issues such as food poverty and farm profitability, she saw a market that was not merely inefficient, but fundamentally skewed.
A food economy tilted away from farmers
At the centre of Sophia’s argument is a simple point: farmers create enormous value but capture very little of it. In some cases, she said, they may receive as little as 1% of the final retail price. Many are struggling to make a sustainable profit at all.
That economic pressure sits alongside a broader set of systemic risks. Climate change is making farming more volatile. Global supply chains are fragile. Cheap food, as the system currently defines it, often comes at the expense of environmental resilience, nutritional quality, and long-term public health.
Her criticism is that the system rewards cheapness over quality and short-term output over sustainability. Farmers are pushed toward intensification because the economics leave them with little room to do otherwise. Meanwhile, larger agribusinesses and retailers are able to use technology, data, and market insight to optimise margins and strengthen their position.
The result is not simply a market failure, but a digital imbalance. The most powerful intelligence sits with the players already capturing the most value.
AI as a tool for rebalancing the system
This is where she believes AI can make a difference. Not as a futuristic add-on or a shiny demonstration project, but as a practical mechanism for redistributing information and influence.
Large companies already use AI to forecast supply and demand, model pricing, and optimise complex trading decisions. Sophia’s argument is that farmers should have access to similar capabilities. If they did, they could better understand what they are being paid relative to others, benchmark contract terms, compare input costs, and identify market opportunities with far greater confidence.
In that model, AI becomes an intelligent layer sitting on top of a farmer-owned data foundation. The real value does not come from AI in isolation, but from combining it with shared data that makes meaningful insights possible.
She sees potential for these tools to help farmers in highly practical ways: identifying grants, measuring environmental impact, spotting relevant schemes, understanding demand signals, and reducing waste by connecting supply more directly to where food is needed.
That matters not just commercially, but socially. She points to the contradiction of food poverty existing alongside large-scale food waste, including cases where crops are simply ploughed back into the soil. Better coordination, informed by better data, could start to close that gap.
Why governance matters as much as technology
Sophia is clear that building a useful platform is only part of the challenge. The structure of ownership matters just as much.
She warns that platforms can easily become extractive when governance is concentrated. In other sectors, digital platforms have created convenience while also centralising control over prices, earnings, and access. Her concern is that agriculture could repeat the same pattern if data platforms are built in ways that benefit intermediaries more than producers.
That is why her preferred model is a cooperatively owned data commons. Farmers, in this vision, would retain sovereignty over their own data, decide who can access it, and potentially benefit financially from the value created through its use.
This is a significant departure from the current model, in which, she argues, large organisations are often able to derive value from farmers’ data without compensating them. A cooperative structure would not only give farmers better tools, but a stake in the platform itself.
For her, data sovereignty is not a secondary issue. It is the condition that determines whether a system genuinely rebalances power or simply digitises the old imbalance.
A practical alternative to AI moonshots
Although she believes the moment is right, Fannon Howell is wary of overcomplicating the solution. She does not frame emerging technologies as the main story. In her view, enormous value can already be unlocked simply by bringing data together, structuring it properly, and making it accessible in useful ways.
She is also sceptical of the current funding bias toward more glamorous agricultural AI applications. Precision tools, sensors, and advanced automation may all have value, but she questions whether they address the deepest problem. Too often, she argues, the industry focuses on extracting more from the land rather than improving the economics for the people working it.
Her priority is more basic and arguably more consequential: make it easier for farmers to farm, report, access support, understand their market position, and get paid more fairly for what they produce.
That may be less eye-catching than some of the sector’s more ambitious technical visions, but it is where she believes the greatest real-world impact lies.
The adoption challenge
The hardest part, she acknowledges, is not the technology. It is adoption.
Farmers are busy, and any successful digital product has to be built around their time, needs, and daily realities. That means starting with clear, immediate use cases rather than trying to build everything at once.
One example she gives is grant discovery: a tool that could identify relevant schemes based on a farm’s profile and surface them automatically would save time and reduce friction straight away. Small, tangible benefits like that are likely to be the path into deeper engagement.
In other words, the long-term vision may be systemic, but the route there has to begin with utility.
A different vision for the food system
Looking ahead, Sophia believes that if AI is applied in ways that genuinely empower farmers, it could help reshape the food system more broadly.
A better balance of information and market access could strengthen local and regional food networks, reduce waste, lower transport emissions, and keep more value in local communities. It could also help make regenerative and nature-friendly farming more economically viable by aligning financial incentives with better environmental outcomes.
She sees a major opportunity in public procurement as well. The UK spends billions each year on food for schools, hospitals, and other public services. Better coordination of supply could open more of that demand to farmers who are currently shut out.
At the policy level, shared data could also help government make better decisions across food, farming, health, and the environment, areas that are deeply connected but often managed in silos. A stronger data foundation, in her view, would make it easier to design more coherent and effective policy.
Building the foundations first
Sophia welcomes the growing interest in AI across government and industry, but she believes there is still too much focus on the most visible applications and not enough on the foundations that make systemic change possible.
Her case is ultimately a simple one. The future of AI in food should not be judged only by technical sophistication. It should be judged by whether it helps create a system that is more profitable for farmers, more resilient for communities, and more sustainable in the long term.
Aterra is still at an early stage, and she is realistic about the scale of the task. No founder or single company will solve this alone. But if the next wave of AI is used to correct market asymmetries rather than deepen them, the technology could do more than optimise the food system. It could help rebalance it.
Watch on Youtube: https://youtu.be/a4fximv6_k8
Listen on Apple Podcasts: