64% of Americans Say AI Is Moving Too Fast — and Farm Data Is the New Battleground

Sixty-four percent of U.S. adults — including 76% of Democrats and 54% of Republicans — believe artificial intelligence technology is developing too quickly, according to a September poll by the Associated Press and NORC. That apprehension ran through nearly every conversation at this year's Big Iron Farm and Construction Show in West Fargo, North Dakota, in mid-September, where Agweek hosted panels on autonomy, weather and precision agriculture.

Farmers make consequential decisions months in advance — seed and fertilizer purchases chief among them. AI promises efficiency and better outcomes. It also raises questions nobody at Big Iron fully answered: whose data is it, how accurate is it, and who can hijack the machine running it.

What can AI actually do for a grower?

Dr. David Mulla, a University of Minnesota professor and executive committee member at the National Institute of Artificial Intelligence in Land, Economy, Agriculture and Forestry, laid out the case in a January webinar. Scouting a thousand-acre field for an early aphid infestation is nearly impossible for one person. Tracking moisture, fertility, disease and insects across that same field, he argued, would take an army.

"Synthesizing all that is kind of difficult for a human," Mulla said. "So increasingly we're seeing artificial intelligence being used to identify patterns, look for hot spots, look for anomalies to help with making better decisions."

His framing is candid, not utopian. "There's always uncertainty in making decisions because we can't know the system completely, but the better information we have, the better our decisions are going to be."

The catch: AI doesn't always have the best data yet.

Whose forecast do you trust?

WDAY meteorologist Lydia Blume described building forecasts from hundreds of ground reports filed by real people across much of North Dakota and portions of Minnesota and South Dakota, cross-checked against detailed patterns and trends. When a forecast shifts, a live meteorologist adjusts it in close consultation with the National Weather Service — person to person.

AI-generated weather content now ships on countless mobile apps. It pulls from a host of sources. But meteorologists stay confident in their jobs, and farmers sometimes shake their heads when the app misses badly.

Lanny Faleide of SatSpot-firm SatShot put a sharper edge on the accuracy problem. "My concern is we're interpolating a lot of data," Faleide said. "But I'm concerned about the accuracy of the data that's coming through, that is now creating assumptions and decisions that maybe are on the edge of danger."

Tommy Grafenauer, a meteorologist with the National Weather Service in Grand Forks, said weather reporting stays safe from AI takeover because of the human component.

"I truly believe that technology can really only take us so far," Grafenauer said. "Any decision that I make where I don't have a relationship with somebody is a worse decision than when I have a relationship with somebody where I understand what they're saying."

Who owns the imagery of your fields?

The recurring panel concern: as AI harvests data to drive decisions, where does that data go, and who owns it?

Shane Williams of TerraClear, whose systems map farm fields to plan rock-picking routes and weed removal, gave the growers' answer directly. The data created on a farm belongs to the farm owner, full stop — including the detailed imagery his company collects in the process.

"We believe there will be a time where you're going to see a value of that data placed in the marketplace, kind of like we saw with carbon credits," Williams said.

His advice is blunt. Any company collecting imagery of your farm should be able to tell you exactly how it keeps that data safe and secure.

Rob Proulx, an NDSU Extension ag tech systems specialist, pointed to the practical gap: most tools are downloaded with one click, and most users never read the fine print on how their data gets used.

Help exists. Ag Data Transparent, a nonprofit working to protect farm data since 2014, reviews agricultural tech contracts and publishes which companies meet the highest standards for transparency, simplicity and trust. Michael Frank at the AgTech Engine in North Dakota said his group is building protections with several startups.

"That's some things we're working on with several startups at the Engine, is how to make an ecosystem around that to protect that, to make sure the security stays with the farmer, with the data, but then also the value stays," Frank said.

Can autonomous equipment be weaponized?

The starkest exchange came on the autonomous equipment panel, featuring Frank, Grand Farms tech advisor Noel Anderson and Jennifer Edney, CEO of Edney Distributing. All three welcome machines that execute programmed work. All three worry about who else could take over the program.

Edney questioned whether regulation can keep pace with fast-moving technology. Anderson said lawmakers must work with stakeholders to identify dangers and build safeguards.

"Ag equipment can be weaponized," Anderson said. "What are the things that need to be done with the components and equipment systems so that someone can't hijack an anhydrous tank and go down Main Street releasing anhydrous?"

Edney called the scenario the stuff of nightmares. Anderson didn't soften it.

"That's a realistic concern, that there are bad actors who would love to do that in a malicious way," he said. "So how do we make sure our systems are designed in a way that will prevent those sorts of bad things from happening?"

The panelists agreed that as automation shifts toward AI making more of its own decisions, safeguards must ensure those decisions are sound, safe and ethical — and that lawmakers have a role in making that happen.

What does this change for growers today? Contract diligence, mainly. The AP-NORC numbers, the Ag Data Transparent reviews and the Big Iron panels all point one direction: the value in farm data is rising, and so is the cost of ignoring who holds it.