Farmers Don't Trust the Government's Advice Anymore — They Trust Each Other, and Increasingly, AI
A 37-farmer study across New Zealand finds trust has quietly migrated away from formal institutions toward peers, the internet, and AI — not because the science got worse, but because it never adapted to how farmers actually want to learn.
Interviews and a focus group with 37 New Zealand dairy and sheep/beef farmers found that informal, digital, and peer-driven information — including AI, used like a search engine — has become farmers' default, while government and research institutions are seen as disconnected and slow. The information age didn't just add new channels; it quietly moved the center of trust, and the abundance of information it created is leaving many farmers more anxious, not more informed.
The question
Who does a farmer actually call when something goes wrong?
For decades, the working assumption in agricultural policy was simple: research institutions and government produce good information, and farmers receive it through an extension system built to carry it downstream. This study, from a team at Lincoln University and the New Zealand Institute for Bioeconomy Science, set out to test whether that assumption still holds by asking 37 real dairy and sheep/beef farmers a blunt set of questions: what sources do you actually use and trust, how do you check whether something is true, and can you find the strategic information you need when a real decision is on the line?
The answer, mapped out in the study's own diagram of farmer information sources, complicates the old picture badly. Formal institutions — government, universities, industry levy bodies — are still in the mix, but they sit at the edge of a much busier network dominated by informal, digital, and interpersonal channels: peers, family, social media, podcasts, and, increasingly, AI.
The method
Twenty-two interviews, one focus group of young farmers, and a literal map of trust
The researchers combined 22 semi-structured interviews (13 sheep and beef farmers, 14 dairy farmers, lasting 45 to 120 minutes each) with a single focus group of 10 young farmers aged 19 to 22, deliberately selected because of their heavier use of digital and emerging information channels. In the focus group, farmers didn't just talk about their information habits — they physically mapped them, drawing themselves at the center of a page and connecting lines out to every source they actually use, with line thickness showing how much they rely on each one.
That mapping exercise is what produced the study's clearest finding: formal institutions cluster on one side of the map, informal and digital sources dominate the other, and the thickest, most confident lines run to peers, the internet, and — for younger farmers specifically — AI tools used the same way they'd use Google.
“It's [AI] good to use it, like you can't miss out on these things or you will get left behind. But you have to make sure that you have an understanding. It's like they don't give you a calculator before they teach you how to do maths. Because you'll get found out if you don't understand something.” — a young farmer (participant YF8), quoted in the study
The finding
Formal institutions lost the popularity contest, but not for the reason you'd expect
It would be easy to read this as farmers losing trust in science. That's not quite what the data shows. Farmers in the study repeatedly said they valued science — they just found it disconnected from their actual working reality. One farmer described research institution output as "complicated stuff"; another said they had "very little idea" what research institutions actually did with their time. The frustration wasn't with the substance of the science, it was with the delivery: too technical, too slow, too detached from a specific farm's specific conditions.
Government fared even worse specifically on the trust dimension — several farmers described a sense that policymakers had "a lot of naivety about what happens on farm," and wanted communication to flow both ways rather than only downward. Industry bodies and processors landed somewhere in the middle: useful for production-specific questions, but farmers described feeling boxed into "sector silos" whenever they tried to use those channels for bigger strategic decisions, like whether to change land use entirely.
The twist
More information didn't make farmers more confident — for many, it made them anxious
The study's most striking finding isn't about which sources won. It's about what winning cost. Nearly every farmer, regardless of age, described real difficulty coping with the sheer volume of available information, much of it visibly contradicting itself. One farmer put it plainly: research exists to support almost any position you already hold, "and then you can go find something that conflicts or counteracts it and they're all peer-reviewed... how do you find the right information for your structure, business, or decision?" Another described the feeling as "an absolute tsunami of information."
This wasn't abstract discomfort — farmers connected it directly to their well-being and decision-making capacity, describing information as landing in the "too-hard basket" and choices getting deferred rather than made. Younger farmers, despite embracing AI and social media more readily, weren't exempt: several described genuine concern about echo chambers, confirmation bias, and outright misinformation circulating in exactly the channels they relied on most.
The fix farmers actually want
What farmers actually want isn't more data — it's someone standing in the middle
The study's most concrete, actionable finding is about farmer-led catchment groups — community organizations tied to local watersheds that blend the structure of a formal institution with the trust of a peer network. Farmers involved in well-run catchment groups reported real increases in how relevant and usable information felt, and several described these groups evolving into something closer to a hub than an information channel: a place for local knowledge-brokering, sure, but also for social connection during hard seasons.
One sheep and beef farmer summed up the mechanism directly: knowledge brokering "at a catchment level or a community level" is simply more likely to be adopted than the same information delivered top-down. The paper frames this as a call for institutions to stop trying to reach farmers directly and instead invest in the trusted intermediaries — catchment groups, respected peer farmers, and farmer influencers — who already sit at the center of the real information network.
This isn't a New Zealand-specific quirk — it's a preview of a challenge every agricultural research and extension system is going to face as trust keeps migrating toward digital, peer, and AI-mediated channels. If institutions keep optimizing for producing more research rather than for how that research actually reaches a working farmer's decision, the gap this study documents will only widen. The paper's core policy recommendation — invest in trusted local intermediaries rather than trying to broadcast past them — is a genuinely exportable idea for extension systems well beyond NZ dairy and sheep country.
Did farmers actually trust AI, or were they just using it?
Mainly younger farmers used it, and even then cautiously — treating it like an advanced search engine rather than an authority, while explicitly worrying about using it without foundational farming knowledge to check its output against.
What exactly is a "catchment group" and why does it matter so much here?
A catchment group is a farmer-led community organization tied to a local watershed, blending formal structure (sometimes with funding) with informal, community-driven trust. The study found these groups were unusually effective at knowledge brokering and were repeatedly named as the kind of trusted intermediary farmers actually want more of.
Does this mean farmers don't respect scientific research?
No — farmers in the study said they valued science highly but found it inaccessible in practice: too technical, too disconnected from their specific farm context, and delivered through channels that don't match how they actually seek information day to day.
Is this pattern specific to New Zealand?
The study's data is NZ-specific (37 dairy and sheep/beef farmers), but it explicitly situates its findings alongside comparable research from Europe and elsewhere, suggesting the shift toward informal, digital, and peer-driven information is a broader trend rather than a New Zealand peculiarity.
Based on the peer-reviewed paper “Bridging Gaps: A Study of Trust and Information Flow in New Zealand Agriculture”. Read the full abstract, key findings, and download the PDF on the paper's own page.