Transformative Principal

Transformative Principal artwork

Practical Ways to Use AI in Your PLC with Kevin Donohue

18 minAudio

Kevin Donohue, a Catholic school leader turned consultant at FACTS Education Solutions, explains how AI can take work off the plate of professional learning communities. Recorded at ISTE 2026, he shows how to use AI for scheduling, agendas, norm-setting and missing expertise, and how to keep student data out of the tools. Donohue leads a session for about 150 teachers and leaders and names five places to plug in AI. His privacy fix is simple: with 24 students in a class, label them A through X so the AI never sees a name, birth date or address, yet can still say which student needs support. What you'll learn Give AI your teachers' schedules and energy levels in the prompt and ask when your PLC should meet. Use AI to write agendas and questions, and to moderate norm-setting so no one teacher steers the conversation. Bring AI in as the missing expert on topics like English language learning, ADHD or dyslexia, with clear guidelines and guardrails. Strip student names, birth dates and addresses from spreadsheets and meeting notes, and use letters A through X instead. Look at student data yourself first, then ask AI to check your own bias in grading and data discussions. Chapters 0:00 Kevin Donohue's path through Catholic school leadership 2:06 Where AI fits in a PLC 3:47 AI as moderator and missing expert 6:34 Making AI another voice at the PLC table 8:00 Protecting student data when you use AI 10:19 Grading versus feedback: where humans stay in 15:13 Leveled readers and differentiation with AI 16:51 One thing principals should do this week Questions this episode answers How can AI support a professional learning community? Donohue sees five places to plug in AI. It can find a meeting time, build agendas and questions, moderate norm-setting, supply missing expertise such as English language learning or dyslexia, and act as another voice at the table. He says it should add to the team, not replace it. How do you protect student privacy when using AI with student data? Donohue says once you give data to an AI, the company has it. He tells teachers to remove names, birth dates and addresses from spreadsheets, agendas and AI meeting notes, and to label students A through X so the AI can still point out who needs support. Should AI grade student work or give feedback? Donohue thinks humans should look at student work first, because that builds relationship and shows gaps, and AI can then check teacher bias. Jethro Jones says AI can grade right-or-wrong work, but feedback needs nuance and should come from the teacher. About Kevin Donohue Kevin Donohue is a former Catholic school teacher and principal in Los Angeles and Boston who now consults for FACTS Education Solutions. Working on this at your school? Get a free preview of a 10-week leadership pathway built from episodes like this one at transformativeprincipal.com/start. Thanks to our Sponsor Broadcast Made Easy who help you start a news broadcast at your school by having everything you need in one easy-to-download place. Check it out today and get your kids creating. We’re proud to be sponsored by IXL. IXL’s comprehensive teaching and learning platform for math, language arts, science, and social studies is accelerating achievement in 95 of the top 100 U.S. school districts. Loved by teachers and backed by independent research from Johns Hopkins University, IXL can help you do the following and more: Simplify and streamline technology Save teachers’ time Reliably meet Tier 1 standards Improve student performance on state assessments 🚀 Ready to see why leading districts trust IXL for their educational needs? Visit IXL.com/BE today to learn more about how IXL can elevate your school or district.

Audio and video play straight from the show's own hosting. PodSubs stores and serves no media.

Transcript

From the show. Tap a time to play from there.

Pete Hall: Welcome to the Transformative Principal. I am so excited to share this interview from ISTE 2026 with you. Really fascinating things happening there. Lots of great people, and over the course of the next couple weeks, you're gonna hear some of these great interviews. Hope you enjoy, and thanks for listening to Transformative Principal Jethro D. Jones: All right, Kevin Donahue at ESTE. you, have a background in Catholic schools, and then go from there. Kevin Donahue: Yeah. So I, I've been a teacher and principal in Catholic schools on both coasts, Los Angeles and Boston. So I entered school

leadership twenty-sixteen, twenty-seventeen with… We had a principal residency program. Mm-hmm. So modeled on the idea of a medical residency, and especially in our Catholic schools, we don't often have the resources to have a full-time VP or AP or And so the program, through generous funders, uh, the Smet Foundation, uh gave five of us a year to be that- assistant principal. Resident principal was our title. And so I spent a year in another school seeing their systems.

We did additional reading and training, and then entered the job pool in the Archdiocese of Los Angeles, the rest is Jethro D. Jones: That's awesome. And how many schools are in the Archdiocese of Los Angeles? Kevin Donahue: So, in Catholic school terms, we are the largest archdiocese in the country. Um, about elementary schools. About another 50 to 60 high schools. So you're talking over 50,000 students- across- Jethro D. Jones: all of LA Yeah, Yeah, that's huge. That's amazing. Yeah. and do you know John Maholio?

yeah. Okay, good. Have you been on his show yet? Kevin Donahue: I have not been on John's show. So I currently, I work for FACTS Education Solutions, and John is actually one of our coaches, so. Oh, is he? Um, yeah, and fantastic, fantastic in both the public school and the Catholic School Jethro D. Jones: just a little plug for his show, Catholic School Leaders Podcast. If anybody's listening and hasn't seen that one, it's a good one, Kevin Donahue: good one. I I, recommend it to I Yeah … especially those in the Catholic school space, but I think increasingly in other spaces too, a lot of what a private school leader has to do is, is falling more and more on public school and charter

school Jethro D. Jones: Really interesting. So, talk to us about, um, you're here presenting about AI in PLCs. Tell me about that. What does that look like? Kevin Donahue: Yeah. So over the last two years, I've been blessed through FACTS to do a lot of presentations on AI, and I think we're now entering that time where it's not, what is AI or "Give me the basics," but it's how do I make this work for me?" Mm-hmm. And I think AI is one of those tools where, especially when it comes to our PLCs, there's an opportunity to take some stuff off of our

plates- Yeah so that we can focus around those, those four big questions and really kind of pursue Jethro D. Jones: Um- So what does that look like day-to-day, though? Because, we're gonna-- We'll address privacy and putting kids' data in there and all that, but what does that look like to you? Kevin Donahue: So we, I was supposed to lead an hour-long session. We had about 150- Oh, nice. teachers, leaders in there, big crowd. Um, of walking I kind of see five opportunities to plug AI in. So kind of first off, especially for summer months, is that idea of planning,

Um, having AI take a look at everyone's schedules, everyone's energy levels, being part of that prompt too, and saying, "Okay, when's the best time for us to meet?" Um, and then what having AI kind of plug in as agenda creator, as question asker, so that when you're coming to it, um, when you're even forming those norms and those root beliefs, which you really should do at the beginning of the year, having AI be the moderator of that conversation so that there's a little less maybe, uh, thumb on the, on the, on the questions that, you know, one teacher's taking the lead teacher's or, something like that.

It kind of frees up the conversation. Jethro D. Jones: Yeah. and that's an interesting role for the, for the AI to take as a moderator. How else do you-- What else is valuable about that Kevin Donahue: So, um, I think where you really see that extension is, is when you're moving into the school year, and I think especially, again, coming from a, a private school background where our PLCs tend to be smaller across the grade levels, where you might not have all the instructional resources of a larger district or school, um, is AI as an expert, right? Or AI as a, as a gap So I-- The, the joke I made during our presentation

was if you have, a, PLC where emotions run a little bit high, maybe you're bringing AI as a moderator that keeps the temperature low. Uh, if you're a PLC where there isn't expertise in English language learning or supporting students with ADHD or dyslexia, maybe AI is that voice around the table so that whatever you-- After your data analysis, after your questions, even during your assessment creation, those conversations are having that kind of expertise brought to the table, you know, through, through the you know prompting with the, r- with the right guidelines and guardrails around that, for sure.

Yeah. Jethro D. Jones: uh, So, ISTE, ASD released, at this conference the Stretch AI. they're gonna release it tomorrow. But it's embargoed till then. But this won't go out till then. Yeah. Uh, so they're-- they released Stretch AI, which is their, their chatbot that allows you to use their resources from ISTE and ASCD to say what best practices look like and things like that, and they've trained it on their stuff. So w- what I see is that for most people,

what the AI has already is good enough. Yeah. Um, and they don't feel the need to go that deep. Is that your impression also, or do you think something like Stretch AI is really that valuable that it'll be used enough for it to be meaningful? Kevin Donahue: it to be meaningful? I think there's a good-- That's a really good question. Um, in this morning's session that I was leading, one of the things we talked about was whatever AI you're using, right, having that project or that folder or that, um, notebook, depending again on Gemini versus Copilot versus, you know, ChatGPT,

um, so that you are at least including your own resources, right? And that might be old agendas, chat conversations, whatever. So I think if there's an opportunity to load in some expertise to make sure those conversations are aligned with broader school goals or district goals, I think that's a real positive. Whether or not that is coming directly from the ISTE ASCD world, it's kind of a different question. Yeah. Um, you know, I think there are a lot of you and schools out there that would be grateful to have those resources as part of, you know, the background

information, right? And this, and this is where like a tool where NotebookLM I think is so valuable is is by ne- being able to choose those sources, you can kind of refine the output and get what you're looking for. But I think for most PLCs, they're, they rightly should be more focused on kind of their students in that aspect. Jethro D. Jones: so when you say being a, a, a, another member of the PLC, what does that look like in practice? Walk me through what that would look like. Are we doing a voice AI? Are we having them be part of the conversation? The AI likes to interrupt a lot, so that may not be the best, but what's your thought there?

Kevin Donahue: So we, we had a great conversation about this. One table mentioned, you know, do you start know, with analysis and then have AI kinda come in? What did we miss? Or do you have AI start the conversation, and then we cr- either critique or add onto or replace what AI is saying? And so I think there's benefits on both sides. Um, we definitely don't wanna do is either cognitively offload all our data analysis, just like our students tend to do, right, um, to the AI. Nor at the same time do we want to, um, kinda question everything, right?

We wanna see it as an add-on, not a replacement for. And so, um, I think kinda when it comes down to it at the end of the day is, is using that voice mode in an agreed-upon kind of norm-centered way in that conversation I think is most helpful. And being thoughtful and considerate of am I providing the right context to the AI in my prompting or in the documents I've uploaded? And am I also bringing the right critical lens when we do get that feedback from the AI? So that's not taking over the meeting, but again, it, it's adding another perspective, possibly reinforcing the skills that we might lack, um, and ultimately

improving student outcomes. Jethro D. Jones: and so let's talk a little bit about the student data and things like that. how do you keep that from being, uploading everything to NotebookLM with all your student information? 'Cause that's not what we do. Kevin Donahue: 100%. And that was important thing that I've tried to bring up repeatedly in this session because, once you give it to the AI, the company's got it, right? Um, a lot of great vendors here, and that's the reality of kind of the age we're, we're living in. They're giving you something for free, and what they're taking from you is your data. Yeah. So, I always

stress go through your, you know, your Excel documents, Word documents, whatever you're including. And again, we need to be careful around agendas. Are we mentioning student names? If we are using an AI note-taker in our meetings, I mentioned Granola during the session, um, are we making sure that that's removing student names as well? 'Cause that's identifiable information. so names, birth dates, addresses. And again, it, it could come down to something as simple as I have 24 students in my class. They're A through Y, A through X, and just kind of plugging that in so that we know student M is Kevin

Donahue, um, but the AI Yeah. Right? And then we can still get some outcomes like, oh, student M really needs extra support with these standards, or maybe we can create an assessment to measure their particular progress over a particular time period. But, we definitely don't wanna submit that kind of personal identifiable information to, to an AI. Jethro D. Jones: Yeah. One of the things that I did when AI first came out, or rather when ChatGPT first came out, is I would, uh, I had a script in Google Sheets that would change everybody's name to something else.

Yeah. And then I just had to remember, map that back to who they were. And I, kept a list so I would know. But, but that really made it a lot easier that I didn't have to worry about putting specific student information in. Unless, of course, somebody figured out who the, who those fake names were and tying them back to the real names. But, little safeguards like that are worth the extra effort to make sure that our students' privacy is protected. But beyond that, there are other things that you can do, like give them, like not usernames and just say this particular student.

just l- give the AI their data only and not any of their personal information. and then you can set up, um, skills and stuff where the AI will know what to do with the information that it has once it has it and you can analyze each student very quickly. Um I had another question that I was gonna ask you about the cognitive offloading. what did you-- where do you sit on the start with the data and human analyze it, or have the AI start with it Kevin Donahue: it? Yeah, that's a great question, and it, it comes up often when I'm doing AI PD just right?

Oh, can AI do my grading for me? That's what I spend so much time on, right? Um, as a principal as a parent, I'm not really the idea of an AI doing all the grading. Um, I think or teachers need to be looking at the data, right? And it's part of building a relationship with students. It's part of guiding them helping them grow, seeing their gaps ourselves. So I, I come down on the, I think humans should look at it first, and then AI is there for those things Um, Right if talking now a little bit about grading, maybe more than a PLC conversation, AI us correct our own

biases, right? Like, we know these kids. We see them every day. Yeah. I've had those kids in class that- Yeah I love a lot. There's no way Johnny wrote this. Yep. And, um, you know, showing preferential treatment or the opposite of that, negative treatment to a particular student for whatever human reason I might have that day, um, AI can look at and say, "Hey, you know, you're being a little harsh here when it comes to the grading." Um, I think that same thing can be happening on PLCs, right? We talked about in our session that when we're looking at the data, sometimes as teachers, we're quick to make excuses like, "Oh yeah, that was the day we had that

school assembly, so that's why our scores are low." Or, "Johnny's having a really tough day that day. That's why he didn't score so well." Um, we might be quick to make that excuse, but if we're bringing in AI as focus on that kind of quantitative side, it can be maybe a little bit more honest before we apply some of that teacher personal interpretation of the data lens. Jethro D. Jones: Now, I am of the pretty strong opinion that it is totally fine for AI to grade, but AI should not be giving feedback. Yeah. that's fair So that's how I would say

that. And the thing is that a grade to me implies right or wrong, you get it or you don't, and that's a binary thing that an AI can do a pretty good job at. Giving feedback, which is more delicate and needs more nuance and needs more, "I understand the kid, and I want to give them good information to help them improve in this area," that to me needs to be the human doing that. What's your take on that? Kevin Donahue: Jethro, I think that's, that's like a great right? And especially if we're looking at something, You know, we've had machines grading

Scantron forms for years. Jethro D. Jones: the, that was the was was gonna give. Like, that w- that would be dumb to have a teacher do that. And an LLM can handle things that are more advanced than a bubble sheet. And so if that's the case, yes, we should take advantage of that technology. I've said for a long time that if an, if a computer can do the job better than a human, that computer definitely should do that job so that the human can do the other jobs that the computer can't do. Kevin Donahue: Yeah …But I think you're right. At the same point, that, that feedback point where we do know the

student, we're seeing kind of what's happening every time. I mean, this this to me a lot of that conversation around standardized testing, right? And I always would tell parents as a school leader, you know, you're seeing your child each and every day, but you should take them to the doctor for a checkup once a year to make sure they're growing properly, they're getting their nutrients and their vitamins," right? As teachers, we're seeing our kids every day, but we want that standardized test to just kind of assess unbiasedly against a national norm how are they doing. Yeah. Um, I think when we're giving that feedback every day, whether it's informal or formal, whether it's formative or just informative,

um, summative, I think it's, it's important that the students or that the student and the teacher are building that relationship together, and that it is kind of going back and forth, right? That the, the student's owning his or her learning just as much as the teacher's looking at that Jethro D. Jones: them, aren't they? Yes, absolutely. And truly, learning is an individualized thing no matter what. Yeah. And as much as we try to make it like, "Here's what we're doing as a class," and doing all this stuff, it really comes down to each individual student has their very own learning path that nobody else can influence. Two twins can be sitting next to each other, same background, same lifestyle, and same teacher, sitting in the same class, they are gonna get different things out of the class.

And once we recognize that, instead instead of thinking that- Just because I say something as the teacher, all the kids are going to get it. Once we recognize that every individual path is-- every learning path is individual, then we have this opportunity to say, "All right, what can we do? What do we need to do as the teachers to help them with this and help them learn and grow?" And that's really where it comes back to the PLC question. There was a point, everybody. It comes back to that, that, the question is: What are we gonna do when these kids get it, and what are we gonna do when they don't

get it? Right. And- Three And- four. Yeah. what is the whole point? And I, I believe that an AI can really help in that arena of customizing things- individualizing it, and making our jobs as teachers a lot easier in that regard, but not take away the human things that we've got to still be doing. Kevin Donahue: that we've got to still be doing. Right. Right. And I think it really comes down to it's another tool in the tool belt for those kind of tier two, tier three, um, adjustments that we might make, right? For those students that need an extra challenge or those students that need an extra support.

you know, I-- It's a teacher pay teachers for free, right? It's, It's, those resources that you maybe have struggled as a teacher to either find or create that aren't coming from your curriculum, that aren't personalized to your kids. Now you've got something that can help you do that. Yeah. Um, starting as A fifth grade teacher, inner city LA. A lot of my students were way below grade level on reading. Most phonics readers are not made for fifth grade boys. You know, Or fifth grade girls. And so I would sit there trying to adapt material to their grade level, and now I think to myself, "Oh my gosh, a tool like Project Read, a

tool like Google Gemini," right? I can-- I could create a reader that would teach the phonic skills that my fifth graders would need, but at their, at their interest level based upon the topics they care about. And I, I… That to me is just a solution that I wish I had, you know, in a time machine fifteen years Jethro D. Jones: Yeah, totally. Yep. I think about myself as a teacher, how much different it would've been if I had access to these tools. would I be better? I hope so. But I don't know. I'm certainly much better at the things I try to do today because of AI, that's for sure. So before I ask my last question, which you know what's coming, what is, How can people get in touch with you?

How can they learn from you? Kevin Donahue: Sure. more than happy to share the presentation. We had sample prompts. We had some work, work time. work reach me via email or, um, LinkedIn, Twitter, all those places. My email is kdonohue, K-D-O-N-O-H-U-E, not, not like Phil. Yeah. at facts, F-A-C-T-S-M-G-T.com, Facts Management. Jethro D. Jones: Okay, excellent. So what is one thing that a principal should do this week to be a transformative principal like you? Kevin Donahue: I think the number one thing I always say is, is, just keep learning.

Yeah. Like, there's so much, especially here at ISTE, but honestly, online podcasts. Jethro D. Jones: Let's go. Kevin Donahue: Yeah. Let's go. YouTube, newsletters. It, it's we are in a golden age for learning, and when you put AI on top of that and the ability to give AI a topic and have it teach You You know, I've had Claude teach me Claude, right? Um, in the session today, "Hey, Gemini, help me do this," and help the teacher do it right there. So, um, just keep learning. Jethro D. Jones: Yep, that's right. I totally support that. Kevin, thank you so much for being part of Transformative Principal today. It was awesome. Thank

Open the transcript file

Hosts and Guests

Hosts

Credit the people on this episode

Be the first to rate this episode

More episodes

All episodes
  1. Can AI Keep Students Safe Without Surveilling Them? | Bella Davis

    27 minAudio

    About This Episode To spot harm on a student's device, a safety tool has to look at what the student does on it. That tension is at the center of this conversation with Bella Davis, founder and CEO of Monarch AI...

    Credit the people on this episode

  2. Stop Owning Their Improvement: Put the Ball in Their Court

    16 minAudio

    A teacher improvement plan often becomes a second job for the principal, who has to follow up on everything. Jethro plays a clip from the My First Million podcast with executive coach Joe Hudson, then shows how to put...

    Credit the people on this episode

  3. Project Paideia: Teens Build with AI Instead of Consuming Slop

    20 minAudio

    Jethro Jones explains Project Paideia, an effort by a group of dads to give teens ages 13 to 18 a better first look at AI: build something that solves a problem instead of consuming AI slop. He separates AI in education...

    Credit the people on this episode

  4. Developing the Leader Within with Christopher Grode

    46 minAudio

    Christopher Grode, Ph.D., retired after 19 years as an Illinois superintendent and wrote The Leader Within on how school leaders grow. He and host Mike Caldwell agree that principal burnout lives in days spent reacting...

    Credit the people on this episode

  5. How Play Builds Belonging and Resilience with Elizabeth Cushing

    27 minAudio

    Elizabeth Cushing, CEO of Playworks, tells Jethro Jones how organized recess builds belonging and emotional resilience, and why it helps student attendance. Playworks is a national nonprofit that has spent 30 years...

    Credit the people on this episode