RAIN — Week of October 19, 2026

Ask Amy

I don't have a tech department where I work. How do I know if the AI tool I'm using for paperwork or notes is actually safe, or if I'm putting people's info at risk?    Back Office Georgia

Back Office Georgia, the fact that you asked means you are already thinking about what you share. Not having an IT department doesn’t mean you are behind. Having one only goes so far. You are still the best judge of what is confidential and what is not.

My advice, first and foremost, is to pay for whatever AI tool you are using. If you are not paying, your data is likely being used to train the model. However, paying for a tool doesn’t make it private. Look for a business or workplace account from a vendor you trust, not a personal one. Does it use your work to train its model? How long is your data kept? Can you delete it? Who else processes it? If you can’t find clear answers, that is a red flag.

This matters most with the notetaker that joins your meetings and the browser extension that summarizes documents. Convenience is not the same as safety.

If confidential information will be discussed, turn the notetaker off. Most of the time you don’t need the real names at all. The tool doesn't need a client’s name and birthdate to fix your grammar. You can get a lot of power out of AI without handing over the names.

If your paperwork touches health records or student files, stop there. Those records come with legal requirements, and some organizations need a formal agreement with the software vendor. A personal account you set up yourself isn't enough.

Back Office Georgia, pick the AI tool you use most and read the details of the privacy policy. If you cannot get clear answers, keep people’s information out of it. That is the safest policy an office without a tech department can have.

Until next week,

Amy

Every column starts with a reader question. Yours can be next: ruralai.org/rain

Stories of the Week

Reservoir Farms Expands to Washington State to Accelerate Physical AI for Tree Fruit, Blueberries, and High-Value Crops

Reservoir will open Reservoir Farms Washington in the Yakima Valley in spring 2027, its fourth on-farm site for rugged physical AI. Partners include the Washington Tree Fruit Research Commission, the North American Blueberry Council and the Tri-City Development Council, which is putting up money, with John Deere committed to $10 million over three years (also covered by AgFunderNews).

Labor runs 50% to 57% of variable orchard costs for Washington apples, and growers say they will not buy machines that have never survived a local season. Startups get free farm access through an associate tier, and WSU's Prosser extension center is in talks to join.

Native News Online Survey: Readers Oppose Data Centers Near Tribal Communities, Demand Tribal Say

In a survey of 115 readers taken October 14 to 18, 75% opposed data centers on or near their tribal communities and 91% said tribes should have enough say to withhold consent outright. Water use, energy demand and environmental impact topped the concerns.

Respondents still named education and research (41%), health care (38%) and language preservation (30%) as real AI benefits, which is the opening for tribally governed projects. With Congress and 36 governorships on the ballot, siting and consultation are now voting issues in Indian Country.

Facing Pushback on Data Centers, Tech Giants are Promising Hosts More Money and Fewer Secrets

An update on the data center backlash covered here recently: Amazon, Google and Microsoft now say they have stopped requiring local officials to sign non-disclosure agreements, and Meta and Amazon have each pledged $1 billion in host-community payments. Google says its local program spending hits $1 billion by the end of 2027 and that it will pay full property taxes on new campuses.

Organizers in rural Montour County, Pennsylvania call the pledges propaganda, but the concessions confirm that small-county permit votes are the bottleneck on AI buildout. Boards negotiating now have more leverage, and more disclosure, than they had a year ago.

Kubota adds M5L tractor to Agtonomy autonomy pilot program

Kubota North America extended its pilot with autonomy software firm Agtonomy to the M5 Low Profile diesel tractor, targeting orchard and nut growers. The driverless system handles spraying, mowing and underplant cultivation, and the M5L's lift capacity covers more orchard jobs, including sanitation mowing, than the narrow-row M5N.

The pitch is retrofitting autonomy onto diesel fleets growers already own, sold and serviced through existing dealers. For specialty crop operations short on skilled labor, that is a far lower cost of entry than a purpose-built robot.

A First Nation community in Canada is using a data center to preserve its heritage

Wiikwemkoong Unceded Territory in rural Ontario installed a garage-sized data center, four server racks cooled by fans, and is pulling its records back from a Microsoft facility 190 miles away in Toronto. It holds health data, corporate records, video of Anishinaabe language speakers, and footage from 700 cameras and a rooftop drone fleet.

Chief Tim Ominika frames it as ownership, access, control and possession of community information, the opposite of AI firms scraping Indigenous knowledge without permission. Delegations from Ontario and Manitoba are touring the 4,000-person community because the model is small enough to copy.

Missouri's AI leader focused staff training and early successes for state innovation

Missouri's AI director describes using Google Vertex AI to process 3,000 Beef and Dairy Producer Grant applications, cutting a five-month review to one month and getting money to farmers faster. The ag-grant example is a genuinely rural hook; the rest is state government AI governance.