Information Technology Committee - Regular Meeting

Good evening. We'll call this meeting of the IT Committee to order. We'll begin with a roll call. Mr. Clark? Here. Mr. Carroll notified me that he would not be attending. Mr. Hassinger? I'm here. Mr. Remillard? Present. And Mr. Robbins is present. We have a quorum. Well, first item on our agenda, as always, is public input. We don't seem to have any Zoom attendees, but if any of us has something that qualifies as public input, not otherwise on the agenda, now would be the time to bring it up. Hearing none, the next item on the agenda This being our first meeting after the start of the fiscal year is the time that we are directed by a charter to reorganize, to select a chair and a vice chair. Bob.

Yeah, I move that we keep everything the way it is.

I second that. All right, I move and second that we keep everything the way it is. That being myself as chair, and if memory serves, Amar as vice chair. All right. Any discussion? Anybody want to change anything? Okay. With no discussion, we'll put it to a vote. Mr. Clark?

Aye.

Mr. Hassinger? Aye. Mr. Remillard? Aye. And Mr. Robbins outside motion carried unanimously. We'll keep on keeping on here. Next agenda item is update from town administrator. I checked with the town administrator yesterday evening and neither he nor the assistant town administrator had anything they wanted to bring to our attention. So the update from town administrator is well, actually nothing. So the next agenda item, and this is mainly why we're meeting tonight, is AI policies. Lucas had distributed a draft AI policy prior to the last meeting, but for whatever reason, we didn't all get that. A copy that has since been provided to all of us, both Bob Carroll sent it around after the meeting, and I included it in tonight's meeting packet. Okay. So let's get into it. Lucas, if you'd like to walk us through what you've come up with, and then we can discuss it from there.

Yeah. So basically, I took, you know, Brookline's, kind of like I said, this policy is modeled and structured on their type of control style. And I kind of pulled in some military style, like the way we do our memos and stuff. So this will help establish how us as the IT committee should review, recommend, and oversee acceptable municipal use of generative AI intelligence tools and similar emerging technologies for Grafton. Now, currently, a lot of people, myself included, and I think Ammar, these tools can help improve productivity and help with drafting notes and summarizations and research and stuff. But, you know, there should be some sort of control on how we do that. Mainly because some people will, maybe Amara could talk to this too, but I've seen people who just, they take notes with AI and then they don't review it. They don't look at it. They're just like, here's the notes. And there's like tons of errors and things, or not, I shouldn't say a ton of errors, but There's errors, there's things that need to be corrected. So, you know, part of using AI is that you review and oversee whatever it is that you generate with your AI tool. You know, and then I went into like the scope of what it, um, this would do definitions and then, you know, the use policy, which has several different things that we can either keep in, add in, remove, um, you know, fairness, non-bias. committee review factors, compliance. The biggest thing was like the responsibilities, because for us, you know, we got the committees, but mainly you're going to have the town employees who are probably going to use this the most. So having the responsibilities of what they do with AI and then different references. And then we have the revision history, which, you know, as we if we adopt this as we do every, you know, if we're going to review it every year, this particular revision history will get longer and we'll, you know, provide whatever updates. And then you got approval on adoption. Went too far.

Okay. Questions or comments?

I read through it and I thought it looked like a really good start and My main question now is to ask, uh, the town administrator, what he thinks about it. Is it coming in the right way to it? Is it the right space? You know, that sort of thing. Um, look pretty good to me. I liked the idea that, uh, it mentioned us a lot.

Yeah.

Yeah, agreed. I think it's a really solid start for this. And I think that a couple of things, you know, I think that we're definitely going to need everybody to weigh in, TA and the board, to weigh in on how the reviews of different tools happen and things like that, right? Make sure that they're okay with it. But obviously that's part of the approval process for any policy. So I think we're covered there. um on the committee review factors um i'd like to see just like a sentence or two maybe at the top explaining what that's what that is right um just just to make it a little bit easier to to read and understand other than that though i thought it was a really solid um policy that you wrote there yeah

Yeah, I think the big question that popped into my head was this assigns the IT committee the role of reviewing AI tools, reviewing and approving. I guess there's nothing wrong with it. It's just I wasn't expecting that because from what Evan had asked for, He appeared to be focused more on just basically advising users on what they can and cannot and should and should not do with AI. And that's part of this, which is entirely appropriate. But I don't know whether... Nothing Evan said to me suggested that he was envisioning that either this committee or anybody else would be reviewing AI tools for appropriateness and basically approving them for use. And it may be that Evan just hadn't thought of that. So that's the one thing that may go beyond what Evan was asking for. Or once Evan gets a look at this, Evan might say, yeah, I think that's a good idea. So I think with whatever changes we end up discussing further tonight, I think this is probably in good enough shape for us to share it with Evan and ask for his opinion, his thoughts. We could end up going back and forth a couple of times to until we get something that everybody's happy with. But yeah, I think that this is generally, I'd say this is a pretty well-written policy. I think it covers everything that I had thought we would want to cover and more.

Cool.

Yeah, just on the... Pass it to him and particularly ask him about this whole idea of the committee having an active role

Yeah, and I would just add that, you know, one thing that came to mind as I was reading that with the IT committee being in the loop is that since we only meet once a month and we don't always meet every month, right? I wouldn't want to get in the way of progress, right? If there is some sort of a tool that the town wanted to adopt and we weren't able to review it in time, that would be my only concern about having us in the loop.

I think that's to be worked out. You can have something like the chairman or his designee can give an interim outlook.

preliminary approval or something like that.

Yeah. Yeah, I think you're right. This is a detail that we can, you know, Evan can weigh in on whether he thinks, whether he wants the committee to be fully in the loop on all proposed use of AI tools or whether he wants to be somewhat more selective about that. It's Probably fair to say that even without a review component of this, pretty much everything that's said about the do's and don'ts for using AI tools would apply across the board.

Yeah. I mean, we could also change it to where it's like we review it once a year, because technology You know, changes, it doesn't change super fast, but it changes consistently. So then that way, like this revision part, you know, we just, we review it. If we need to make appropriate changes based on emerging technologies, we make an update and then we just relabel it as, you know, revision 1.2 with the date we did it. And we could do it like once a year versus reviewing it every time they want to use a new type of AI.

I think that's a pretty good point, is that the policy, given the state of the industry and how technology changes, the policy itself ought to be reviewed. I think once a year is an appropriate interval to review the policy. Aside from reviewing the policy, there is a part of us that's about reviewing and approving the AI tools. And that's something that is much less predictable on a time basis. It's whenever someone proposes using a new AI tool, that's when this policy as written would bring us into... review the tool itself and make some recommendations or approvals or disapproval or whatever. But the policy itself, yeah, absolutely. It should be, I couldn't say reviewed and updated as often as necessary, but generally speaking, I think a year is probably appropriate.

Okay. I will write something up.

Yeah, and I think that we want to be careful not to avoid having perfect get in the way of progress, right? I think having a policy is better than having no policy, right? Even if the policy isn't 100% perfect, then we can always revise it as we go.

That sounds almost exactly like what I heard. Who was it? When I went to that MMA webinar, one of the participants, one of the towns that was participating there made a point of saying, you know, they've updated it about six times and they're still not convinced it's done.

Yeah. Well, the field keeps changing. Um, that's why I prefer to take what we've got, maybe even with side comments about possible changes, but this month getting it moving to, um, to find out if we're in the right hemisphere. And, you know, he can feed us back. We only meet once a month. There's no law that says that it has to be that way. You know, that's as needed.

Yes, that's absolutely true. So We've discussed a few, I would call them minor tweaks, which Lucas is taking notes on as we speak. And so, you know, with anything else, you know, any further discussion we have on this, Lucas can revise the proposal basically right after this meeting based on what we've discussed. And then without delay, we can share that with Evan. seek his input on that and follow up on that at our next meeting.

Right. I look at Lucas sending it to you and you're sending it to, you're looking at it and sending it to Evan. So moved.

I kind of wrote something up already. Do you guys want to hear it? Go for it. Yeah, go for it. So this is kind of what I had. The IT committee shall review this policy at least once each calendar year and recommend updates or revisions to, I put Evan's name, but I'll put town administrator and select board as appropriate to ensure the policy remains current with advances in AI technology, cybersecurity, any updates to applicable laws, municipal operations and industry best practices. Sounds good.

Did you say at least as well?

At least once each calendar year.

But it could be more. The way you're saying it, it could be more.

It could be more, but we're at least just required once a year. If there was... They came out with quantum AI where it completely changed everything. We could review it and make updates to it, but... you know, part of this is at least once a year we do it.

Yeah, saying at least once a year gives us the flexibility to review the policy as often as we see a need to do so. Yep. Or for that matter, if someone in the town administration thinks If someone else identifies a need to update the policy, they can bring that to us and we can review it. So a review, other than the obligatory once a year, could be initiated by any relevant party that it comes to us.

Cool. All right. I'll make those updates and send it to send it to you, David.

Yep. And I will, yeah, I will include it as I've done this more often with the charter review committee. Cause we do a fair amount of discussion of documents, you know, introduction or discussion of documents at the meeting, but I'll, I'll, I'll call, I'll call this the work product result of this meeting. And what I typically do with the Charter Review Committee is I post a, after the meeting, I'll post a supplemental meeting packet with any documents that the committee had discussed and maybe updated during the meeting. So that way we have it on record as a public document for everybody to see.

I think somebody moved to go ahead with this, didn't they? I'll second that.

All right. Moved and seconded. Any discussion of the motion? Hearing none, Mr. Clark? Aye. Mr. Hassinger? Aye. Mr. Remlard? Aye. And Mr. Robbins outside. Motion carried unanimously. Nice. We had something accomplished, and it didn't take us a tremendous amount of time to do it. What are we doing here? We'll move this on to the next step, which will be to share it with Evan for his feedback. Okay, so was there anything else that we needed to discuss on the topic of AI policies? I think this is really the thing we were shooting for. All right, moving on to the agenda then. We have the minutes of the June 18th meeting to review.

Move that we approve them with any corrections we discuss. Seconded.

Moved and seconded. Is there any discussion? We'll call for a vote. Amar?

Aye.

Bob? All right. This time I can say Bob without any ambiguity on this meeting. Lucas. Aye. And Dave votes aye. Motion carried unanimously. The minutes are approved. The next agenda item is possible executive session. I know of nothing that requires an executive session. So the last agenda item would be adjournment.

Move we adjourn. I second.

Moved and seconded. We adjourn. I'd like to say all in favor, but I actually have to call the names. Amar.

Aye.

Lucas. Aye. Bob. Aye. And Dave Boatseye. Motion carried unanimously. We are adjourned.