How to Put AI to Work for Your Association

Naylor’s AI Solutions Manager shares how associations can use AI for knowledge handoffs, member questions, accuracy checks, and more. 

As AI Solutions Manager at Naylor Association Solutions, my job is to help people take an idea for how AI could make their work easier and turn it into something that works for more than one person. When people bring me those ideas, they rarely ask for something flashy. Most of the time, someone is spending hours on work that doesn’t require their expertise, and they want that time back. 

I think that’s the right way for associations to think about AI, as well. Association staff and volunteers are already stretched across member service, committees, education, events, advocacy, and governance, and there is almost always more work than there are hours to do it. There’s definitely interest in using AI; in the 2026 Association Benchmarking Report, 55% of respondents rated AI tools as very or extremely valuable. 

The most useful thing AI can do is take on some of the work around those responsibilities, so the people who carry them have more time for the parts that need a human being’s judgment and experience. 

Start with the Work, Not the Tool 

If you’re not sure where AI fits in your association, look for the places where people spend time moving, summarizing, reformatting, finding, or comparing information. That might be the person who compiles committee discussions every month, the staff member who pulls chapter reports together for the board packet, or whoever gets asked the same procedural question five times a week. Those tasks are usually better starting points than asking which AI tool your association should buy. 

The tools are also changing too fast to build a plan around them. Something that looks like the best option today can be replaced by a new feature six weeks from now, and I’ve watched that happen over and over in the past year. The problems your staff deal with every day will still be there, so start by understanding the work and choose the technology afterward. 

Get the Knowledge Out of Janet’s Head: Capturing Institutional Knowledge 

Rather than relying on long-time staff or volunteers to retain processes and procedures, easily create a permanent record of what they know. Here’s how that process looks. 

Every organization has a Janet. She knows everything about a particular program or process, including why it’s done a certain way, what happened the last time someone tried to change it, and which exceptions never made it into the written procedure. None of that is written down, so when Janet goes on vacation for two weeks, or leaves for good, the knowledge goes with her. 

Associations deal with this more than most because so much knowledge lives with volunteers. Committee chairs, board members, and subject-matter experts rotate in and out of their roles, and every handoff is another chance for something important to get lost. 

AI makes it much easier to capture what Janet knows. At Naylor, when one of our internal projects moved to a new owner, we recorded a conversation where the outgoing owner walked through why the project started, what it was solving, how the processes worked, why decisions were made, and where things left off. We gave that transcript to Claude, had it create a handoff document, and saved the document in a shared drive connected to Claude, so anyone with a question about the project can ask it in plain language and get the answer with the context behind it. 

Any association can do the same thing the next time a staff member changes roles or a committee chair hands things off. Next time, try this: 

1. Record your handoff meeting. Most meeting tools, including Zoom and Microsoft Teams, can transcribe a call automatically, so you don’t need anything new to get started. Let Janet know you’re recording.  

2. Use the meeting transcript and ask AI to create a formal handoff document that can be used in future, similar instances. 

3. Before you rely on the finished document, have Janet read it over and correct anything the AI got wrong or left out. 

4. Save the document in a shared space so it’s readily accessible. 

Help Staff Find Answers for Members Faster 

One of the safest places to start with AI isn’t having it talk to your members directly. It’s helping the people who talk to your members find the right answers faster. 

If a staff member gets a question about a certification requirement or an old board decision, AI can help them find the right internal document, pull out the relevant part, and point them back to the source. Ask for sources every time and open them before you answer, because AI can sound completely sure of itself while being wrong, and it will sometimes cite a source that doesn’t exist. 

We’re using the same idea at Naylor behind the scenes. One of our business development managers receives every inquiry that comes in through a form meant for new business, and plenty of what comes in isn’t new business. Address changes, accounting questions, and messages meant for other departments all land with him, so he spends time figuring out where each one should go. That’s a coordinator’s job, and it takes time away from managing his team. 

We’re building an inbound coordinator that reads each request and routes it based on criteria we’ve defined: these requests go to this department, these types of questions go to that person, and these need a human to look at them. A person is still copied on everything, and when something goes to the wrong place, we review what happened and adjust the instructions so the routing improves over time. 

How to Check AI’s Work for Accuracy 

Before you rely on anything AI produces, ask it how it verified its work, then ask it to make a separate pass looking for its own mistakes. Here’s how that played out in our Gainesville, Florida office.  

Our IT team needed a count of all the technology equipment in the building, and it turned out nobody had ever kept one. I work out of that office, but the people who could identify the equipment were in Canada. Our technology team suggested that instead of flying someone in for a day of counting, I take photos and let Claude build the inventory. 

Here’s how it worked: 

1. I labeled each room with a sticky note, took about 130 photos, and put them in a folder connected to Claude along with a map of the office.  

2. I didn’t have a plan going in, so I asked Claude to help me figure out the approach. It asked me questions first, including what the final file should look like, what was in scope, and how my labeling system worked, and then it built the inventory from the photos. 

3. The most important part came after that. I asked Claude how it had verified its accuracy, and when it went back and checked, it found mistakes.  

4. Then I asked it to make a separate pass that compared each row against its photo and looked for errors instead of confirming its earlier work, and that caught even more. 

The whole project took about three hours, with no flight, no hotel, and a photo record IT can go back to. If accuracy matters, ask the AI how it verified its work, then ask it to look for its own mistakes. 

Decide Which Member Interactions Should Stay Human 

You can decide upfront which tasks are fit for AI and which need the expertise and finesse of a human.  

I’m a member of a small association, and recently I reached out to them because I needed advice about something affecting someone important to me. By the time I sent that email, I had already searched online, read everything I could find, and asked ChatGPT for more information. What I wanted was a person’s perspective, someone with experience who could tell me what they’d seen. 

The response I got back had one or two sentences acknowledging what I was going through, and then several paragraphs that were clearly written by ChatGPT. The information was helpful, but it wasn’t what I had gone to the association for. 

Associations have something AI can’t offer, which is people with real experience who can share what they’ve lived through. An association is a living library of experiences, not a large language model of words, and as more of the world goes digital, that human connection matters more. 

So, when you’re deciding where AI belongs, ask whether the hard part of the task is the work around the expertise or the expertise itself. If someone spends 30 minutes hunting for a policy before answering a member, help them find it faster. If a member comes to you for judgment, perspective, or care, that’s the part to keep human. 

Build a Culture Where Staff Can Experiment With AI 

Remember to make the process of AI experimentation a safe space.  

The most important thing an association needs for AI adoption is curiosity, and people are only curious when they feel safe trying something new. At Naylor, we hold a weekly AI brainstorming meeting where anyone in the company can share what they’ve tried, and afterward, I send a company-wide summary. Some of the examples are serious business tools, and some aren’t work-related at all, like the “Pretentious Pour-Over Coffee” app one colleague built for himself over a weekend. Both kinds of suggestions help, because once people see a coworker try something, they start wondering what AI could do for them. 

Your association doesn’t need a program like ours to get the same effect. Give one curious person on staff permission to spend time learning, and set aside five minutes in a staff meeting you already have for someone to share something they tried. It also helps to be clear that not every experiment has to work. Even the experiments that don’t work teach people something, and that’s knowledge you’d otherwise pay for in a webinar. 

One guardrail before anyone starts: if your staff will work with member data, use a business plan instead of personal accounts. Depending on the tool and its settings, personal accounts may use your conversations to train AI models, and they don’t come with the admin controls and data protections a business plan includes. Your members’ data is worth the cost. 

Start Inside 

I still think about that email from my association. If someone there had used AI to find the information faster, then written back to me themselves, I would have gotten both the facts and the personal expertise I was looking for. My advice is to begin there. Use AI for the work inside your association, like handoffs, reports, and finding information so your staff and volunteers have more time for the people you serve. 

If you want somewhere to start this week, find your Janet and put an hour on her calendar. 

To learn more about AI adoption among associations, download the 2026 Association Benchmarking Report.