AI Might Finally Change Food & Ag. The Strategic Questions Start Now.

By Mary Shelman & AJ Shelman. Mary Shelman is the founder of Shelman Group, a boutique strategy consulting firm specializing in the food and ag industry, and the former Director of Harvard Business School’s Agribusiness Program. AJ Shelman is a partner at Shelman Group. The authors can be reached at mary@shelman.co. 

Why Food & Ag Has Been So Hard to Change

Innovation is different in Food & Ag. Over the last 40+ years, the industry has seen wave after wave of hyped innovation — from precision agriculture and digital marketplaces to alternative proteins, vertical farming, blockchain traceability, and online grocery. Many of these have created value in specific pockets. Few have produced truly transformational change across the system. Four structural challenges explain why:

• Complexity creates uncertainty. Food & Ag is the only major industry with a value chain bookended by two highly complex biological systems: the farm and the human body. At one end, weather, soil, genetics, water, pests, disease, and management interact in ways that are hard to predict. At the other end, consumer behavior is shaped by taste, price, convenience, biology, habit, and psychology. The result is a system where nearly every important decision is made under uncertainty.
  

• Behavior is sticky at both ends. Farmers rarely change practices quickly because the risk is real, the capital is committed, and the windows for action are narrow. Consumers rarely change what they eat because food choices are deeply habitual. Decades of nutrition education, sustainability messaging, and food innovation have produced fewer mass-market behavior changes than expected.
  

• Information does not flow well. Data exists across the value chain, but it is often trapped in disconnected systems: farm records, packer reports, ERP systems, retailer portals, spreadsheets, PDFs, and institutional knowledge. Fragmentation and transactional relationships make it difficult for information to move from farm to processor to brand to retailer to consumer in a usable way.
  

• Financial incentives and capital flow are weak. The industry’s relentless focus on low-cost food often leads to antagonist relationships and prevents value from moving to the parts of the system where better decisions could create the most improvement.

Artificial intelligence is the latest hyped innovation to arrive in Food & Ag. The question is whether it will follow the pattern of past technologies, useful in specific pockets but ultimately less transformative than promised, or whether it can deliver real system-level change. We think AI could be the rare technology that delivers transformational change in Food & Ag because it directly addresses the four structural challenges that have limited the impact of so many innovations before it.

AI’s potential impact is easier to see when mapped directly against the four structural challenges:

• Complexity creates uncertainty → AI improves decision making. Food & Ag is full of decisions made under uncertainty. A grower is predicting weather, pest pressure, labor needs, crop development, yield, quality, and market timing. A food company is predicting demand, consumer response, ingredient availability, price elasticity, and product performance. A retailer is predicting inventory, freshness, shrink, substitutions, and shopper behavior. AI does not make biology simple, but it does give the industry a better way to work through the complexity by connecting genetics, environment, management, quality, cost, and demand into more usable decision systems.
  

• Behavior is sticky at both ends → AI lowers the friction of change. Farmers often resist new practices because the risks are concrete and the benefits are uncertain. AI can help translate complex data into specific, timely recommendations: what to change, when to change it, and why the risk is worth taking. Consumers are similarly hard to move because food choices are habitual and emotional. AI may not need to persuade them in the traditional way—advertising, education, brand claims, nudges. Personalized nutrition and agent-enabled meal planning and shopping can make different choices easier by optimizing around taste, price, convenience, health goals, and household preferences. In other words, AI can change behavior not by asking people to care more, but by making the better decision the easier decision.
  

• Information does not flow well → AI increases the value of connected data. For years, the industry has had plenty of data, but much of it has been trapped in disconnected systems and unusable formats. AI raises the value of that data because better models require better inputs. That creates new incentives to collect, clean, structure, and share information across farms, processors, brands, retailers, and consumers. Over time, the companies that make their data AI-readable and connect it across the value chain will be able to see and act on system-level opportunities that others miss.
  

• Financial incentives and capital flow are weak → AI creates new financial incentives. AI could also change incentives by making value creation across the chain easier to see and share. A retailer’s AI system may reduce waste and improve freshness by changing how growers harvest, packers sort, or distributors move product. A processor’s forecasting model may create value only if suppliers share better production data. A grower’s AI-enabled quality improvements may help a brand or retailer win with consumers. In those cases, the economic benefit of better decisions does not sit neatly at one level of the chain. That creates a stronger reason for financial incentives and capital to flow across the value chain, so the players whose actions improve the system can participate in the value their decisions create.
  

That is why AI deserves attention. Its importance is not just that it can make today’s tasks faster or cheaper, although it can. Its larger potential is that it could remove some of the barriers that have kept Food & Ag from changing at scale. If AI helps the industry make better decisions under uncertainty, makes change easier for farmers and consumers, and creates new incentives for data to flow and AI to be deployed across the chain, then its impact could extend well beyond productivity gains. It could begin to reshape the system itself.

AI’s Real Impact Is Strategic Advantage, Not Short-Term ROI

AI adoption across the Food & Ag sector is starting where most technologies start: with practical tools that make existing work faster, cheaper, or easier. Companies will use AI to reduce overhead, improve labor productivity, speed up analysis, automate routine tasks, and tighten forecasting. In a low-margin industry, those gains matter. But they are only the first step.

The most important AI applications will not be the ones with the cleanest standalone ROI. They will be the ones tied most directly to how companies create and defend their fundamental competitive advantage. For an input company, that might mean using AI to discover better traits or crop protection products faster than competitors. For a grower, it might mean using AI to make better labor, irrigation, harvest, and quality decisions within narrow seasonal windows. For a food company, it might mean developing products faster, reformulating around cost or supply shocks, or predicting consumer response with greater accuracy. For a retailer, it might mean using data and AI to create more personalized shopping experiences — tailoring recommendations, promotions, meal ideas, and product substitutions to each household’s preferences, budget, health goals, and buying patterns.

The most powerful examples are already moving beyond generic productivity. Inari is using AI-driven genomic modeling and gene editing to design seed traits more quickly and precisely. Enko is using machine learning and structure-based modeling to discover new crop protection products. Nestlé has used AI to mine consumer and trend data, generate product concepts, and compress early concept-to-prototype timelines from roughly three months to three weeks. Mondelez has used machine learning to refine recipes by balancing taste, aroma, ingredient cost, nutrition, and environmental impact, including in the development of Gluten Free Golden Oreos. Walmart is using AI with sensor and supply-chain data to improve visibility into freshness, inventory, and product movement. On the consumer side, ZOE is using AI, microbiome testing, and biometric data to personalize nutrition recommendations.

While AI is likely to create winners and losers in the near to medium term, in the long run, many of these AI-driven advantages will erode. As the technology becomes cheaper, more capable, and more widely deployed, today’s breakthroughs will become tomorrow’s table stakes. The lasting advantage will come less from having access to AI itself and more from how well a company learns, how effectively it integrates data, how quickly it turns insight into action, and how deeply AI is embedded in the decisions that matter.

Ultimately, the biggest disruptions — and the most durable advantages — may come when companies move beyond applying AI to optimize the current system and start redesigning entire systems around data and AI. The goal will not simply be to make individual functions more efficient, but to connect decisions across every stage of the value chain in ways that allow companies to move faster, see more clearly, and make faster and better decisions than competitors. At that point, the strategic question becomes less about which company has the best individual tool and more about who controls the data, relationships, and decision points that allow AI to work at scale.

Walmart is probably the clearest example of a company positioned to build this kind of AI-enabled food system. It already sits at the point where consumer demand, supplier relationships, logistics, inventory, and pricing all come together. Now it is taking steps to make that system more data-rich and AI-readable, using RFID, ambient IoT, and large-scale data platforms to create better visibility into inventory, freshness, temperature, shelf life, and product movement through the supply chain. Walmart is already connecting downstream demand signals with upstream supply-chain data, and the implications go well beyond store execution. For a retailer with that much demand visibility, better data improves not only internal execution, it changes the ability to influence how the upstream system behaves. Walmart can use that visibility to coordinate more of the system around its own data and decision rules — ranking suppliers with greater precision, setting higher expectations for data sharing and responsiveness, and steering volume toward the partners that best fit its AI-enabled operating model. 

This is where the strategic implications become much larger than any individual AI use case. Companies that can vertically or virtually control multiple stages of the value chain will be able to use AI to make better decisions than players that lack the same level of access, visibility, and data. For suppliers in that chain, this could create a new basis of competition: those that can plug into AI-readable systems may become preferred partners, while those that cannot may be pushed toward lower-value channels. Over time, that could shift Food & Ag toward tighter coordination, more proprietary supply chains, and a sharper divide between companies that help shape AI-enabled systems and those that are forced to operate inside systems designed by someone else.

The pattern is likely to unfold in waves. First, AI will improve productivity and reduce costs. Then it will reshape competitive advantage by changing how companies make decisions. Over time, the bigger shift may come as parts of the food system are redesigned around data, AI, and new forms of coordination.

Don’t Wait for AI to Become Obvious

For leaders, the mistake is treating AI as tactical until its strategic impact is impossible to ignore. By then, competitors, customers, retailers, and technology platforms may already be making decisions that reshape the market around them. AI cannot sit off to the side as a technology project. It belongs in the core strategy discussion: how the business wins, where power may shift, and what capabilities will matter next.

That means looking beyond individual tools and asking how AI could change the competitive landscape around the business. What happens if customers use AI to rethink sourcing and supplier selection? What happens if competitors use AI to lower their cost position or improve service levels? What happens if retailers require more data transparency from suppliers? What happens if consumers begin delegating food decisions to AI agents? These questions are not about whether a single use case clears an ROI hurdle. They are about how the rules of competition and the systems that we operate within could change.

The starting point is to identify the decisions that matter most to competitive position — and where AI could most change how the company wins. For some companies, that may be product development. For others, it may be production efficiency, forecasting, procurement, customer targeting, logistics, or access to strategic channels. The right AI agenda should begin with the places where better decisions would most improve how the company wins.

That requires leadership teams to do three things. First, align around how AI could affect the basis of competition in their sector. Second, play forward scenarios across the value chain, including how customers, competitors, suppliers, retailers, and technology platforms might use AI. Third, start learning now by assessing data readiness, building internal capability, and experimenting in the areas most connected to competitive advantage.

Food & Ag has resisted transformation for good reasons: it is biological, fragmented, low-margin, behaviorally sticky, and capital-constrained. Those same characteristics are why AI could matter. It may start with productivity and cost reduction, but the larger impact will come if AI changes how advantage is built, how value chains coordinate, and how parts of the food system are redesigned.

Waiting for perfect clarity is the wrong move. By the time the impact is obvious, the most important decisions about data access, partnerships, channel power, and system design may already have been made.

ABOUT MARY SHELMAN

Mary Shelman is a driving force in global agribusiness, guiding organizations toward a sustainable and innovative future. Formerly the director of Harvard Business School’s Agribusiness Program, she now leads Shelman Group, providing strategic consulting to the ag and food value chain. Shelman leverages her broad industry knowledge and unique insights into global trends to help clients craft and execute winning strategies. A Kentucky native where she still owns a farm, Shelman holds a Harvard MBA and brings a passion for improving the food system from farm to fork. She actively contributes to the agribusiness community by serving on company boards – like the Women in Agribusiness (WIA) Advisory Board -- judging agtech competitions and frequently speaking on the future of food and ag at industry events. She has been involved with WIA since its founding, co-chairing the inaugural WIA Summit in New Orleans in 2012. 

Shelman Group is a boutique strategy consulting firm dedicated to helping clients create and execute winning strategies in the complex Food & Agribusiness industry. We combine deep expertise in strategy with a broad view & unique insights into the global Food & Agribusiness value chain to help clients succeed.