AI Adoption on US Farms Splits by Size, Not Age, as Agribusinesses Race to Build Tools

How AI Is Actually Showing Up on American Farms

Artificial intelligence is spreading across American farms, but not in the way many observers expected. A new analysis by Marc Rosenbohm, a senior grain and oilseeds analyst at Terrain, published in Bushels' 2026 State of the Farm, finds that AI use cuts across all farmer age groups. The familiar assumption that younger producers would lead adoption does not hold. Instead, the dividing line runs by farm size: operations below roughly 5,000 acres are more hesitant, while larger farms are moving faster.

The reasons for that hesitation are practical. Rosenbohm points to time constraints and uncertainty about what AI would actually deliver. Where farmers are using AI today, it is mostly general-purpose tools — ChatGPT, Google Gemini and Microsoft Copilot — applied to office work such as drafting and editing documents. Among mid-to-upper-sized farms, use is starting to extend into business and financial analysis, though Rosenbohm cautions that the depth of that use is not well understood yet. He also flags data silos as a barrier, noting that AI can help connect insights currently trapped in separate systems.

Equipment and input companies are pushing the technology further into the field. Ryan Schaefer, vice president of New Holland North America, says the company is using AI to improve machine technology and data analytics, including crop quality assessment as material reaches the combine. Jeff Cecil, president of North American crop protection for Syngenta, describes using AI to evaluate what worked and what did not in past formulations, then to identify promising paths for new ones. That process, he says, can narrow more than 150,000 molecules down to 50,000 for further evaluation — and Syngenta launched its first fully AI-driven molecule in Brazil last year.

Jim Knuth, senior vice president at Farm Credit Services of America, frames the shift as inevitable rather than radical, comparing it to the smartphone transition farmers have already absorbed. In his view, AI will soon be a routine way to drive efficiency in farm businesses, not a novelty.

Why Farm Size, Not Age, Predicts AI Adoption

Farm Size Is the Real Adoption Divide

The headline finding — that AI use is not an age story — matters because it undercuts a common assumption in agtech marketing. If younger farmers were the natural early adopters, vendors could target succession-age producers and wait for generational turnover. Instead, Rosenbohm's data suggests adoption tracks scale, which is a harder problem to solve. Larger farms have more staff, more data and more complex decisions where AI can plausibly save time or money. Farms under 5,000 acres often run lean, and Rosenbohm cites time constraints and unclear benefit as the main brakes. That is a classic adoption gap: the technology is available, but the perceived return on the time required to learn it is not yet obvious.

General Tools Beat Ag-Specific AI — For Now

The finding that farmers lean on ChatGPT, Gemini and Copilot more than agriculture-specific AI is significant for the competitive landscape. It means general-purpose platforms are capturing farm workflows before dedicated agtech vendors have locked them in. Office tasks — drafting, editing, basic analysis — are the entry point. If those habits harden, ag-specific tools will have to integrate with, or clearly outperform, the general platforms farmers already know. The data-silo problem Rosenbohm identifies is the opening: AI that connects machinery data, agronomy records and financials into one view solves a problem general chatbots cannot, and that is where specialist vendors can still win.

Syngenta and New Holland Show Where the Value Is

Syngenta's molecule-screening example is the clearest commercial proof point in the story. Narrowing 150,000-plus molecules to 50,000 for evaluation compresses an early-stage research step that traditionally consumes years and significant cost. The launch of a fully AI-driven molecule in Brazil — a major agricultural market — signals the approach has moved past pilot stage. For Syngenta, the payoff is a faster pipeline and potentially lower R&D cost per commercialized product. For competitors in crop protection, it raises the bar on discovery speed. New Holland's use of AI for crop quality assessment at the combine points to a different value pool: hardware-embedded intelligence that improves machine performance and gives the manufacturer a data-driven services story to sell alongside equipment.

What This Means for the Farm Economy

For context, farm margins have been pressured in recent years by input costs and commodity price volatility, which makes efficiency-focused technology more attractive — but also raises the bar for proving ROI. The pattern described here echoes broader enterprise AI adoption: general tools first, specialized applications later, with scale determining who moves fastest. If larger farms capture the efficiency gains first, the gap between large and small operations could widen on cost per bushel, not just on technology. That is a structural question for lenders like Farm Credit Services of America, whose exposure depends on the profitability of farms across the size spectrum, not just the largest ones.

Winners and Losers

Likely near-term winners include large farms already experimenting with AI-driven financial and business analysis, general AI platforms capturing farm office workflows, and input and equipment companies like Syngenta and New Holland that can translate AI into measurable product advantages. Those at risk of falling behind are smaller farms that delay adoption and the ag-specific software vendors that fail to solve the data-silo problem before general tools become the default.

What Farmers and Agribusinesses Should Do Next

The single most important insight: on US farms, AI adoption is a scale story, not a generational one — and the farms under 5,000 acres that hesitate now risk a widening efficiency gap with larger operations.

For farmers and farm managers:

• If your operation is under 5,000 acres and you have not started, begin with one office task you already do — drafting contracts, summarizing agronomy notes, editing documents — using a general tool like ChatGPT or Copilot. Rosenbohm's data shows this is where most current farm AI use sits, so the learning curve is low.
• If you are on a mid-to-upper-sized farm, the next step identified in the survey is business and financial analysis. Test AI on one recurring decision, such as comparing input costs or reviewing marketing scenarios, and track whether it saves measurable time.
• Address data silos before scaling. Rosenbohm specifically flags fragmented data as a blocker; connecting machinery, agronomy and financial records is the prerequisite for AI to deliver anything beyond document drafting.

For agribusinesses and agtech vendors:

• General-purpose AI tools are already inside farm offices. Ag-specific products must either integrate with them or demonstrate a clear advantage on tasks general tools cannot handle — starting with data integration across farm systems.
• Syngenta's path — using AI to cut 150,000-plus molecules to 50,000 for evaluation and launching a fully AI-driven molecule in Brazil — is the benchmark for crop-protection R&D speed. Competitors should expect pipeline timelines to be judged against it.
• Equipment makers like New Holland are embedding AI in machinery for crop quality and data analytics. Dealers and service networks should prepare to support and explain these features, since adoption will depend on farmers understanding the value.

Metrics worth watching: the next Bushels State of the Farm update for whether sub-5,000-acre adoption closes the gap; Syngenta's pipeline announcements following its Brazil molecule launch; and whether general AI platforms or ag-specific vendors capture the next wave of farm business-analysis use.