Start With the Problem, Not the AI Tool with Victoria Lansdown of North Star AI Enablement

Victoria Lansdown is the founder of North Star AI Enablement, helping schools, Edtech companies, and nonprofits adopt AI in practical, human-centered ways. A former educator, she has led AI-for-educators programs at Google and contributed to UNESCO’s AI competency framework discussions.

💡 5 Things You’ll Learn in This Episode:

1. Why effective AI adoption starts with a problem, not a tool.
2. How educators can balance AI experimentation with meaningful guardrails.
3. Why AI policies need practical guidance and not just a policy document.
4. How schools and organizations can measure AI’s actual value and impact.
5. What specialized AI agents could mean for the future of Edtech.

✨ Episode Highlights:
[00:02:56] Rethinking computer science education with generative AI
[00:05:32]  AI as an equalizer: The equity questions around AI access
[00:08:18] Welcoming educator skepticism and moving beyond the “pro-AI vs. anti-AI” binary
[00:21:05] Why successful AI adoption starts with a problem, not a tool
[00:25:50] Building AI policies around both governance and implementation
[00:34:07] Creating psychological safety for responsible AI experimentation
[00:42:30] Starting small to build AI confidence and creativity
[00:49:41] The shift from general-purpose chatbots to specialized AI agents and experiences

😎 Stay updated with Edtech Insiders! 

Follow us on our podcast, newsletter & LinkedIn here.

🎉 Presenting Sponsor/s:

Tuck Advisors was founded by entrepreneurs who built and sold their own companies, frustrated by other m and a firms, they created the one they wished they could have hired but couldn't find. One who understands what matters to founders and whose North Star KPI is the percentage of deals closed. If you're thinking of selling your ed tech company or buying one contact Tuck advisors now.

Every year, K-12 districts and higher ed institutions spend over half a trillion dollars—but most sales teams miss the signals. Starbridge tracks early signs like board minutes, budget drafts, and strategic plans, then helps you turn them into personalized outreach—fast. Win the deal before it hits the RFP stage. That’s how top edtech teams stay ahead.

This season of Edtech Insiders is brought to you by Cooley LLP. Cooley is the go-to law firm for education and edtech innovators, offering industry-informed counsel across the 'pre-K to gray' spectrum. With a multidisciplinary approach and a powerful edtech ecosystem, Cooley helps shape the future of education.

Innovation in preK to gray learning is powered by exceptional people. For over 15 years, EdTech companies of all sizes and stages have trusted HireEducation to find the talent that drives impact. When specific skills and experiences are mission-critical, HireEducation is a partner that delivers. Offering permanent, fractional, and executive recruitment, HireEducation knows the go-to-market talent you need. Learn more at HireEdu.com.

Transcript

Copy Transcript Copied

[00:00:00] Alex Sarlin: Innovation in pre-K to grade learning is powered by exceptional people. For over 15 years, Edtech companies of all sizes and stages have trusted Higher Education to find the talent that drives impact. When specific skills and experiences are mission-critical, Higher Education is a partner that delivers.

Offering permanent, fractional, and executive recruitment, Higher Education knows the go-to-market talent you need. Learn more at hireedu.com. That's H-I-R-E-E-D-U.com. 

[00:00:33] Victoria Lansdown: I try to do different things. I try to find a little video that I like on YouTube as a tutorial and just try things out because just find a short one or find one that you're comfortable with and then try it out.

You know, embodied practice is the best way to create those long-term myelinated pathways is what neuroeducation calls them. And so just, you know, experimenting with the tools and seeing what the limitations are. I think that will just naturally inform future decisions, but at the same time, help encourage that optimism and creativity and excitement.

[00:01:06] Alex Sarlin: Welcome to Edtech Insiders, the top podcast covering the education technology industry. From funding rounds to impact to AI developments across early childhood, K-12, higher ed, and work, you'll find it all here at Edtech Insiders. 

[00:01:21] Ben Kornell: Remember to subscribe to the pod, check out our newsletter, and also our event calendar.

And to go deeper, check out Edtech Insiders Plus, where you can get premium content, access to our WhatsApp channel, early access to events, and back-channel insights from Alex and Ben. Hope you enjoy today's pod.

[00:01:46] Alex Sarlin: Welcome to Edtech Insiders. We have an amazing guest with us today on Edtech Insiders. We're speaking to Victoria Lansdown, MBA, who is the founder of North Star AI Enablement, which helps schools, Edtech companies, and nonprofits adopt AI in practical, human-centered ways. She's a former educator. She has built AI for educators programs at Google, and she's contributed to the UNESCO 2024 AI competency frameworks discussions.

She also serves on the Pearson's AI Strategic Council. Victoria Lansdown, welcome to Edtech Insiders. 

[00:02:21] Victoria Lansdown: Thank you so much for having me, Alex. It's a pleasure to be here. 

[00:02:24] Alex Sarlin: It's a pleasure to speak with you as well. So one of the projects we just mentioned, which is really interesting, is that you led Google's initiative integrating generative AI into computer science courses, specifically for historically Black colleges and universities and Hispanic-serving institutions.

Really interesting project, and I wanna start by asking you about that project, and then we'll zoom out and talk more about some of what you do with North Star. But what made this project unique? What instigated the project, and what was it really about as you got it out into the world? 

[00:02:56] Victoria Lansdown: Yeah, great question.

This was a really fun and unique project. So this project actually originally took place back in 2024, and I was the cross-course program lead as a contractor at Google through Synergist, and we worked with faculty from these HBCUs and HSIs across several North American universities to create the first round of a series of gen AI integrated computer science courses, and these courses are actually available online now for anyone to use at teachCSwithAI.org.

And one of our key faculty partners was James Prather, and he's a leader in AI in computer science education research. And so the collaborative approach that we took with faculty, including James, was what made this project unique. So we were all asking how AI could actually change the way computer science is taught in higher education from, like, a fundamental level.

And the question that kept coming up was, what should the students still be able to do independently, and where should AI allow them to do something more ambitious? And so where we took that was in the introductory computer science courses, we decided that we still wanted to have, like, a final assessment, almost like a pen and paper assessment, where the student had to demonstrate foundational understanding without any use of AI.

So that ended up being like debugging code, but not necessarily writing code, because nowadays AI can do most of the code writing. But in the project-based software engineering courses, the final project was designed so that students were expected to use AI and accomplish substantially more in the same amount of time.

So that kind of reflects the environment that they're likely to s- to encounter professionally 

[00:04:39] Alex Sarlin: That's really interesting. One of the things that's interesting about this project as well is this idea that it's designed specifically for students, for institutions that have underrepresented populations.

And I think one of the things I wanted to ask you about is the equity concern with AI. I think there's been a feeling over the last few years that how do we make sure that AI is not distributed sort of unequally between populations? It is a very powerful technology. It's useful in many different workplace settings, and if it's only taught to certain types of students with certain types of access to cultural capital, then you might exacerbate inequality.

I'm curious if that was one of the motivations behind this set of courses designed specifically for HBCUs and HSIs, and also sort of how you see that progressing in the time since. Do you think that there is still an equity issue that we should be paying attention to in terms of who has access to AI education?

[00:05:32] Victoria Lansdown: Yeah, that's a big question and something that definitely was important as we created these courses. The equity piece was part of the discussion, and that's why we sort of went towards HBCUs and HSIs. So the goal was making sure that these students were actually developing experience using AI inside authentic disciplinary work to help bridge that gap and that divide.

And what we hoped students would leave with was for them to understand, like, where AI fits into their own technical work, like coming from their own backgrounds, coming from the future that they may be stepping into, knowing where they still need to show independent competence and how to use it and extend what they're capable of producing with all the barriers they may be facing.

But that equity playing out in education nowadays is also an interesting question, and I kind of like the idea of AI as an equalizer, not a disruptor. 

[00:06:26] Alex Sarlin: Yeah. 

[00:06:26] Victoria Lansdown: I've heard that somewhere. I don't remember who first said that, but I didn't coin that. But there is an opportunity to use this technology to create opportunity, like create personalized solutions that maybe help fill in some gaps of practice or reinforce key learnings.

Uh, and that can help the equity piece. But I also understand this sensitivity and hesitation around, like, technology being student facing. And this is a big question around, like, how AI can continue to create opportunity and be equitable when that may not be in its original design. Uh, we need to just be careful on the information that it's created on and modeled after.

And there is implicit bias in current data that some of these models are trained on. So just being aware of that and having knowledge about that is key starting out. 

[00:07:13] Alex Sarlin: Yeah, I agree. It's been interesting to watch that discussion sort of start to shift over time. But I think we're still in a, a moment where we don't know what's gonna happen in the future in terms of who has access, and it's important to be very aware of creating opportunities for all sorts of different populations.

You mentioned the sort of-- one of the unique features of this project was the connection to faculty and building it alongside higher education faculty. And I think one of the downstream effects of a course like this is that you're working with students and with faculty members in universities to help distribute these courses and so to support them in actually doing the courses.

So you've worked with a number of different faculty members at different institutions, probably, I imagine, all along the spectrum from sort of very, very excited and positive and can't wait to use AI in their computer science teaching to very skeptical and concerned and not sure exactly how it fits in.

I'm curious what approaches you've used to help s-support people along all sides of that spectrum, because honestly, you know, there's cases to be made for every reaction to AI at this particular moment. How do you support faculty members no matter where they're coming from? 

[00:08:18] Victoria Lansdown: Yes, absolutely. So the first thing that we did was welcome the skepticism.

A lot of the skepticism was completely reasonable, and faculty were asking questions like, "Are students still learning the fundamentals?" Or, "How will I know what they actually understand?" Or, "Does this make cheating easier?" These are really valid questions, so we tried to start with the learning objective instead of the AI tool that we're working with.

And if the objective is independent mastery of a concept, maybe AI shouldn't be part of that assessment or part of that course. But if the objective is more applied, like say being able to build a larger software product using current professional tools, then AI actually seems to fit well in the objective itself, and we can start to create opportunities that way.

So I think this is where education gets stuck in this false binary of pro-AI versus anti-AI. 

[00:09:09] Alex Sarlin: Yes. 

[00:09:10] Victoria Lansdown: And the much more useful question we ended up asking or kind of moving the conversation towards was, what should this student be capable of at the end? So we started with small changes, learning about data literacy, learning about how to use AI responsibly, governance, having these key pieces of all courses and creating a starting point for AI use policies, and then being able to adapt and modularize the content as we went.

[00:09:37] Alex Sarlin: Tell me more about the modularized content, because that feels like an interesting approach to supporting faculty along all sides of the adoption spectrum. You know, they might be able to do a little bit and take one module and, and incorporate it into their existing course. They might be able to lean in further and take more or test it out or try it in the context of the different types of assessments and learning goals of the course.

How does modularization support the ability to incorporate or at least test out AI in lots of different contexts? 

[00:10:05] Victoria Lansdown: Yeah. Modularizing the content actually gave faculty lots of examples and different activities that they could take and run with, and that actually also helped the project expand across over a dozen different campuses.

It's for the same reason. So they helped-- This kind of idea of modularizing content helped because faculty could see what AI integration might look like in their own course and then adapt it to where their students were in the semester. So that also helped with the equity piece for HBCUs and HSIs, like we were talking about, or these underrepresented groups in these institutions because these activities in modularized content is meant to be adapted to either large or small class sizes with varying backgrounds.

There's a lot of flexibility in there, and there's a lot of instruction. There are educator resources, there are teacher guides, there's teacher notes in almost every activity and document that's included in the course materials, and there's guidelines and recommendations and resources for them to learn more, like links to the latest research we have at this time.

So there's a lot for them to work with and take what they like and leave what they don't. 

[00:11:10] Alex Sarlin: You're mentioning this sort of idea of this false binary, and I've been wrestling with this a lot recently, this idea of there's either pro-AI or anti-AI, and there's sort of nothing in the middle. And it feels like there's a lot of different interesting approaches to, as you say, acknowledge and appreciate the really meaningful skeptical thoughts that people have as they're thinking about this, about assessment, about cognitive offloading or academic integrity, to be cognizant that these are real issues.

They are not solved issues, but at the same time, not turn your back entirely on this entire system. And if you're teaching computer science at the university level, not incorporating AI might not be the way to prepare people for a career in the field. It's a really nuanced way to look at it. And one of the things you've mentioned is the role of sort of personal interest projects in, uh, helping faculty sort of gather where AI might be useful for their exact class or for their own use cases.

I'm curious how that played a role in helping these dozen campuses incorporate AI into their computer science curriculum. 

[00:12:10] Victoria Lansdown: Yeah, that's a great point. I think one thing I want to start with is there is an opportunity to use things like personal interest projects to be flexible. But the overall point here is that the only way I've seen to move forward is to continue to reevaluate the curriculum, adapt, be flexible, kind of humanize the experience, recognize that we're all learning as we go.

This is fairly new technology, and recognizing that every year we need to reevaluate again. We need to recognize that workflows are changing, the capabilities of the technology are changing. How are we ensuring data privacy as these tools change? What kind of questions should we be aware of? What should we be asking so we can be informed when these changes occur?

But to go back to your question about the personal interest projects, I actually first came across that when working with Neural Education, which is a Washington-based nonprofit, and they look at the research on AI's impact on attention, memory, and cognition. So it's been really interesting to work with them, kind of just hear their story, hear the research they've been doing.

Of course, none of this is really substantial yet because we're building it as we go. We, we haven't had a lot of time to create really significant findings with control groups and everything that we need to create something substantial, but it's in the works. And one thing I want to mention there is having that research being conducted and knowing that these questions are starting to be looked at has helped some faculty and has helped some of this skepticism Because we know these questions need to be answered from a fundamental research level with objective data.

So knowing that that's starting to happen is reassuring in a way. But until we get those results, you know, again, we need to be flexible. So going back to this idea of these personal interest projects, neural education offers these with some higher education institutions or some K-12 institutions that are open to it, that allow the student to choose something that they're really passionate about and something they've never learned before, and try to use AI to solve that problem.

So they're encouraged to use AI in a safe environment and with guardrails and with a policy shared with them that they can either contribute to or, you know, just be informed of. And so within this system or this environment or this class where they're permitted to use AI, we're meant to have open conversations from both sides.

How is the teacher using AI to either create versions of this assignment or maybe in some form of the assessment? It-- that's a very careful question. Or on the student side, how is the student planning on using AI? Or what questions did the student ask their AI tool as they went? How did they check their sources?

How did they know that the output was reliable? What did they do with it? How did they adapt? How did they use critical thinking and implement these transferable skills? And then how can they see that in their industry or in their career? 

[00:14:53] Alex Sarlin: It's really interesting, and the idea of a personal interest project creates a sort of passion-based safe space that's outside of the traditional curriculum, outside of standards, or outside of a higher education curriculum to allow students to really dive into it.

You can create the guardrails, you can create practice opportunities and make sense of it in an environment where the stakes are a little bit lower, but where the passion and relevance may be quite high. I think it's a really powerful technique. Outside of this particular project, you do really interesting work.

You sort of play a really unusual and I think relatively unique role in sort of being an interesting bridge between this sort of educator perspective and AI enablement in various ways. You founded North Star AI Enablement to help schools, universities, Edtech companies, and nonprofits adopt AI in, as you say, practical, human-centered ways.

And this can happen in many different ways in any of these organizations, right? It could be back end, it could be totally administrative, internal. It could be for educators to be creating lesson plans or assessments or content. It could be for Edtech companies to be creating content or AI tooling. Tell us a little bit about how you help institutions understand their AI needs and what lens you bring as an educator to the table to help them understand how they can incorporate AI in a safe, responsible, human-centered way.

[00:16:09] Victoria Lansdown: I love this question. I wanna come back to the educator lens at the end, but I wanna share a bit of context first. So North Star helps educational organizations or these groups, as you referenced, build the infrastructure around AI adoption. And one thing that's intentionally different about North Star that you mentioned is that we don't just work in education, we also support industry organizations with workforce AI adoption and upskilling because we see that work feeding back into the education side.

So if the workplace is changing, educational programming and career readiness curriculum have to change with it too. So that can definitely mean looking at strategy and governance, or redesigning workflows, or training in L&D, or even creating custom internal agents and automation for educators or workforce environments that allow for that, and implementing support mechanisms and measuring whether the investment is actually creating value.

So my team has both the technical and enablement sides, so we help build something that also helps an organization actually adopt it. And to tackle these challenges, North Star often works with Edtech companies as a fractional AI team. And my strategy co-lead, Matthew Campbell, a former head of K-12 digital learning at McGraw Hill for over a decade, and I have this technical team of automation experts that I referenced as well as L&D experts, so we can actually support AI tool and agent development along with educator product feedback and workflow automation.

So as a former educator, I bring that lens in both sides. I have this experience in tech, I have this original background in education, and I have this passion for both. So I think having that same passion and range of experience in both sides of the company really helps to bridge that gap, and that's what allows us to focus on that enablement piece.

And one way we do that is through partnering with Ednology and Ednology experts because they're looking at the larger ecosystem around schools. So they provide kind of a one-stop shop for Edtech products, expertise, implementation, and even funding So North Star brings kind of AI enablement into that ecosystem.

[00:18:17] Alex Sarlin: It's really interesting. And let's circle back to the educator lens because I think, you know, this is something we talk a lot about on the podcast, how, you know, this is a big ecosystem. There's a lot of different players making sure that we are really working together and that not siloed, and that we can have that educator thinking and product thinking and business thinking and funding thinking and, and enablement and tech thinking all sort of happening in the same place.

When you put your educator hat on, what does that lend to the work? 

[00:18:42] Victoria Lansdown: Absolutely. So I think from the educator side, I try to help inform leaders evaluate these AI products and help us continually reevaluate our strategies around how educators are trained. So knowing, you know, boots on the ground, what it's like, knowing educators come with limited time available to them, knowing they have a lot of questions that are very valid, and they want to be heard.

Asking what guardrails institutions can put around AI use with that in mind. You know, how can educators be part of the discussion? And then longer term, I am hopeful for what this means for different levels of support, for different educator personalized solutions, different workflows that we can adapt regularly at a scalable model.

And I think that that same lesson applies at the organizational level, so not just in education. So a successful AI demo doesn't necessarily prove that the product will integrate into like a district or reduce workflow or meet governance requirements or create any measurable value once it's deployed.

Right. So at North Star, we evaluate the full implementation system, and this is where the evidence and need for data and research comes in. So North Star continues to look for partnerships to fulfill that gap from a variety of impacting goals. 

[00:19:56] Alex Sarlin: Yeah. Implementation is such an important part of the Edtech ecosystem, and I think it's often undervalued because it's downstream of some of the business decisions, but it is so key to what actually happens on the ground and whether you're compliant, whether you're actually making a difference for students, whether you're getting the dosage in, whether you're supporting teachers and educators in the training.

I mean, there's so many different aspects to getting an education technology product from a demo, as you say, to actually having impact on students, a positive impact on learning. You need people to really shepherd it through the entire process. So it's really, it's important work, and you have all sorts of amazing people who are contributing to that, but sort of aligning them all and having all the perspectives in place can be difficult.

As you work with different organizations of different sizes, of different sort of comfort levels with AI, I'm curious what you've seen that separates organizations that are sort of really figuring out how to stack all the pieces together and successfully incorporating Edtech or AI-enabled Edtech versus those that where there are still gaps and there are still sort of places that the pieces don't quite fit together.

Is there any patterns you've noticed between those who are sort of making it work and those who are still wrestling with it? 

[00:21:05] Victoria Lansdown: Yes, we've been taking a close look at this so we can implement the same ideas. First off, the successful ones start with a problem, not a tool or a product license. Mm-hmm. So I see organizations buy Copilot or Claude or another platform and then ask everybody to find a use for it, and I try to start my client conversations by reversing that.

So I ask us to find operational or instructional problems first. That might be like an inefficient workflow or inconsistent governance or a professional learning gap or what have you. But the successful organizations that are adopting AI tend to provide the ongoing enablement rather than just an introductory training here and there.

[00:21:44] Alex Sarlin: Yeah. 

[00:21:45] Victoria Lansdown: To offer this at a scalable level, one of the models that we built for Ednology starts with the workshop, but then continues with like monthly PD, office hours, asynchronous support, and creating a living prompt library based on what educators are actually doing, and impact reporting along the way.

So I think the last thing I want to mention there is organizations doing this well measure value rather than adoption. Mm-hmm. So they look at did staff get meaningful time back, or was QA just pushed onto someone else down the line? Those are the enablement measures that show real value, and tracking those allows us to reevaluate often so we can continue to adapt.

[00:22:24] Alex Sarlin: Powerful, uh, principles, right? Start with a problem, not a tool or a solution, measure value, and, and continuously think about professional learning not as a one-time kickoff, but consistently create a sort of community and many, many touch points and opportunities for professional learning so that as the technology continues to be rolled out, there's lots of places where people can get support and help and feedback on what they're doing.

I think those are powerful principles. One of the things that you mentioned that I want to double down on, I think it's incredibly interesting and really relevant to what many people are doing in the Edtech space, is that we're at this very strange moment, I think, in technology history, where seventh grade students and college presidents and administrators at large districts and school principals and educators are all learning at the same time how to use this technology and what are the potential, what are the workflows that are helpful, what are the risks, what are the guardrails.

We're all learning together. And I think one of the aspects of humanizing AI that you've mentioned in your work is being transparent in some cases about that we're all learning together and not trying to create a situation where everybody acts like, "Oh, we have everything perfectly in place, and it's only the students who have to learn.

It's only the educators who have to learn." Just this concept of sort of everybody learning this at the same time and being a little more transparent about the fact that we're learning, we're all wrestling with it. I'd love to hear you talk more about that because I think that's so key to this moment.

[00:23:45] Victoria Lansdown: Yeah, I think that feeds really well into the last question too, around like what are successful organizations doing, and I-- that is a key element, is being humble and recognizing we're all new, and there is almost no way to keep up. So I would just say there are a lot of resources I keep up with to try to stay on top of this.

I look at those in a variety of contexts. I look at them as an educator, I look at them as a business owner, I look at them from the tech environment that I'm in. You know, I try to consider a lot of questions like that so I can be informed from a variety of angles. But I see the same questions coming up.

You know, everybody seems to be overwhelmed by these changes. You know, we went from generative AI to agents, to integrations, to automations, to now we have all these connectors available to us, and it's really helping, but it's a lot and it's happening so quickly, and it's very hard to keep up. And so that-- I think just recognizing that that is a great opportunity to work together and collaborate.

So that's North Star's model, is to continue to look for partnerships and collaborate and come from this angle of being humble and wanting to learn and wanting to create something new together because, of course, when we all work together, we can create something better and more impactful than we can alone.

So that is what I've seen be a successful angle across the board. 

[00:24:59] Alex Sarlin: Yeah. There's some factors built into both work culture where you have, you know, people higher up in a hierarchy sort of needing to project confidence and authority, and in school culture where you have educators often wanting to not appear to be wrestling with anything or seem like they're in complete control of, of a classroom or of, of everything they're doing.

I feel like this is an interesting moment to sort of maybe shake up some of those cultural norms and, and embrace the uncertainty and the fact that we're all learning together. I'm curious, you know, if you have any specific examples of when you've seen that work in a classroom or in a district with any of the institutions you've worked with where people are starting to sort of open up their playbook and say, "This is what I'm trying to figure out.

Maybe we can work together on it," or, "I'm gonna try this, but I'm not sure it's gonna work because I've never done it before and we're gonna have to all work together." I'm curious what that actually looks like in practice. 

[00:25:50] Victoria Lansdown: I do have an example, actually. So one I wanna reference is the AI use policy work we did with a New York school district, Central Islip, through collaborating with educator Alex Luciano.

And the model we used had two layers because we had this open conversation around acknowledging, like, the policy that we're creating is not the endpoint. So we tried to adapt as we went, and we wanted to have these open conversations as we moved forward, and we had two layers. One was governance, and so we looked at concerns around privacy, academic integrity, data handling, acceptable use, and student use versus employee use, and what approved tools we have to work with.

And then layer two is implementation, so we looked at providing role-specific guidance, teacher checklists, classroom norms, examples, decision trees, staff walkthroughs, and FAQ documents. And that's how I packaged policy offerings for Ednology as well, because we've seen that as a pretty scalable model that works to have a recognition that the policy is not the endpoint.

Let's be vocal about how this is something we're trying out. We're not sure if it's gonna work. That needs to be built into the design. We need to continue to reevaluate and reassess if something's working year over year, maybe in every monthly PD session, just having a conversation around what's been working in the classroom and what do we need to adapt 

[00:27:08] Alex Sarlin: Exactly.

Yeah, and just having that sort of learning mindset, that growth mindset of, you know, we're trying this, and let's see what's working and what's not, rather than feeling a sort of cultural incentive to just whitewash and say, "Yeah, it's going well," or to say, "We have the policy in place, and that's all we need."

So speaking of policy, you've mentioned that one of the things that you do with North Star AI Enablement, and that I think you think a lot about, is AI use policies for school districts, and there's many components of them, right? Com- governance, compliance, data privacy, training, you know, professional learning, who uses the AI, who has access to it.

There's so many different aspects of it, especially when you're trying to really dig into what responsible AI use looks like. So tell us about what a flexible, scalable AI policy actually looks like that educators can use and trust and actually feel like it actually guides their work rather than just being a sort of piece of paper somewhere that they know exists.

[00:28:03] Victoria Lansdown: Yeah, great question. I would say one thing that we've seen work is not just having the policy, but normally it comes with three parts. So one is a standard policy that, again, is open to in- interpretation and open to being adapted, and just having that space and recognition that whenever you are using this policy, like reporting back to leadership to say what's working and what's not working, and have those communication lines open in both directions.

But the policy comes with a checklist and a FAQ document or like teacher guide. So that is kind of where I'm going with this, where that has been what makes the policy look successful or actually be valuable in practice, because there's not just this document, there is also a checklist for teachers to easily reference and see, can I ask myself these questions around the applicability of this policy, or how can I evaluate its use?

Or how can I-- what are some things I need to be aware of, or what do I even need to be asking to even know that this policy is even helping me or not? You know, it's just guidelines around the practice of it, and then the FAQ document or teacher resource, or that third element that we have is often really long, and it's sometimes called an appendix.

So we have it-- an example of that on my website, the North Star AI enablement website, and it's again, really comprehensive. It gives prompts, it gives ways to iterate with AI tools. It gives many different use cases for different grade levels. It gives, again, questions to ask. It has-- and it gives resources for the teacher or educator to continue their learning, and we try to update those regularly as well.

So it's a three-part offering, but diving a little bit deeper into the policy itself to answer your question around like what makes the policy successful, we tend to ask questions that we see come up because educators and staff need more operational guidance on things like which tools are actually approved or what information can go into them.

What information about my students can I be putting into this, if any? When is human review mandatory, or can AI be used for feedback or grading support? Or what documentation is needed about how I'm using these tools, and what changes when AI is student-facing versus staff-facing? Right. So North Star in practice mostly Looks at the back end, like staff-facing applications of AI because it's just a safer environment, and that's where the technology is now.

But where schools are asking for more student-facing work, we just are a little more careful and sensitive to those discussions and allowing for flexibility there, but just being extra mindful of the data privacy concerns with 

[00:30:42] Alex Sarlin: that. That makes a lot of sense. One aspect of the use policies that, that we've thought about a lot at Edtech Insiders, and I'd love to ask you about, we just did some research and, and put out a piece in our newsletter about how teacher enthusiasm for AI has actually been decreasing over time.

The-- if you look at teacher surveys, people have been less positive and optimistic and sort of excited about the technology and more concerned about the downstream effects in, in all sorts of different ways. And when you dig deeper into that, some of what-- there's a lot of different potential reasons for that, but one thing you start to see is that a lot of teachers don't feel like there are consistent policies, or they don't feel like they're getting enough professional learning, and they feel like they're sort of out on a limb on their own with this complex technology, with all of these factors.

I mean, you've named probably 20 different factors you have to think about when you're implementing any AI tool, and the teachers feel like, "Well, I'm on my own trying to figure this out, and if I get it wrong, there's risk." I'm curious how you've seen that dynamic play out in your work. Do you feel like the responsible AI use policies and the professional learning sort of excite teachers and enable them to lean in and use the technology?

Or does it feel like, okay, there's this large, complex, rigid framework, and I'm just gonna go back to the way I was teaching before because it's not worth doing? How do you see it playing out? 

[00:31:58] Victoria Lansdown: I have definitely seen both sides, and I hesitate to say a number or a percent that goes in either direction 'cause it's changed every year, but I absolutely see this drop in optimism from educators.

And there's a couple things that come to mind here. One is I know Neural Education hosts these cognitive institutes, and they're meant to be collaborative working sessions for educators to all come together and look at their AI use policies that are offered in their schools and talk about whether this is kind of a top-down or bottom-up approach and whether their input is welcome or not.

I do think, of course, this is kind of an easy answer, but a lot of it is with the delivery and the presentation of the policy and the design and to say, you know, to what extent is this collaborative. At the same time, I think that- Good governance can actually encourage responsible experimentation. So if people know the boundaries, I think that they have some kind of psychological safety to experiment inside of them.

But it depends obviously on the resources that are available. We are all strapped for-- A lot of educational institutions are strapped for resources. So I think that just having, again, a collaborative approach to the policies tends to acknowledge high consequence uses like grading or disciplinary action or placement or student support decisions.

Those are really high consequence uses, and so those are not the same use cases as using AI to brainstorm a lesson. So just being aware of the differences and having that as part of the discussion and then again adapting as we go. But I definitely just wanna reiterate that I see the optimism dropping, and I, I understand the concerns.

I wanna hear more from educators about, you know, what the concerns are. I, I appreciate more of that information coming to light because that is how we move forward and just Wanting more input from educators with, again, those boots on the ground, like in the classrooms. What do you think-- They know their students, and their students can often speak for themselves.

You know, if they ask the student, "Is this working for you?" The student can often say, you know, yes or no. And just having that advocacy, I think, as part of the discussion and part of the criteria and policies is what will help us move forward. 

[00:34:07] Alex Sarlin: That makes sense. I think I like that concept you named of psychological safety and sort of creating tiered use cases.

This idea of, you know, there are high-stakes use cases that have major downstream consequences, and then there may be some that are lower stakes or more room for experimentation, and creating space for that psychological safety and that sort of inspiration and excitement, and teachers actually trying to do activities or lessons that they've always wanted to do but they've never had time to plan, or they've never been able to get all the resources in place, and AI en- enables that.

Like, creating room for the positive aspects and psychological safety around it so that there's no removing or reducing the risk there seems like a really positive aspect. I have one more question about this that I'm curious about, and then I want to ask about the UNESCO framework. But one of the things that has struck me a lot recently is that there's this idea that educators, you name all of the different sort of compliance mechanisms and checklists and things that, that educators need to do.

I'm curious how you would recommend that Edtech companies who are actually trying to deliver meaningful instruction and sell their tools to schools, how they can contribute to the positive aspects of this. You know, how can they create their own materials or create their own training in a way that can feed into the best practices of how it's being used in schools, so that when a school incorporates a, an Edtech tool, there is continuous professional learning, or there are aspects of it that maybe feed directly into their use policy.

I'm curious what role you think Edtech companies might be able to play in supporting schools in going down these paths. 

[00:35:40] Victoria Lansdown: Yeah, that's a great question. I do think there are a few elements of things that I spoke about that I kind of want to call back to that Edtech companies can replicate. One is modularizing the content.

Yeah. So I do see a few opportunities for some of these companies to continue. Some of them are offering many different types of activities based on different age levels and interests and subject matter and the works. I think continuing to dive into that and offering additional ways to modularize the content, or at least providing recommendations or suggestions for a few sample prompts for educators to, say, for example, work with this activity and then find a way to adapt it to their students.

You know, giving examples of ways to safely adapt it and expand and, yeah, again, provide creativity and encourage, like, responsible use, but in a creative way. I think there's a lot of opportunity to continue to offer resources around that and guidelines and guidance around that. So I'd say that's one that Edtech companies can definitely expand on.

Another is Kind of offering similar resources to what I referenced as part of their package of materials. So, I'm thinking of a few Edtech leaders. I think just referencing whenever you have a set of curriculum, noting which tools can be grounded in that curriculum. Pearson has a smart lesson generator that's grounded in Global Scale of English standards and globally recognized standards for English language learning, and recognizing that there are some of these tools that are source grounded and are safe to use, having those as a list of resources for teachers to explore.

Or another element, another direction I'm thinking of taking this is there are so many platforms out there like Magic School AI, Khan Academy, you know, there are these tech companies or like tech angles that have support mechanisms available or present opportunities for Edtech companies to collaborate and then give recommendations using those platforms wherever possible.

So just knowing what resources are available to teachers, giving suggestions for not only how to initially adapt or continue the material provided, but then to continue to have conversations with either AI tools or with their communities to continue to expand those. 

[00:37:52] Alex Sarlin: Yeah, that's helpful advice. Let's talk about the UNESCO framework for a bit, because you contributed to the discussions on, on the UN framework for, for AI education, and we're coming to this really interesting moment.

Obviously, you know, every individual organization sort of n- is developing in state and district, you know, all different levels are trying to develop their own AI policies, but we're also seeing some very large organizations and nonprofits start to contribute and try to figure out how might we contribute to this conversation in a meaningful way, add our own expertise.

I'm curious what those discussions were like and what was sort of most exciting about working with UNESCO on their framework. 

[00:38:30] Victoria Lansdown: Yeah, absolutely. So just to give some more context about how I was involved. So in 2024, in September, I attended the UNESCO Digital Learning Week, which took place at the UN headquarters in Paris, and that's where I participated in the discussions when they first launched the, again, 2024 initial AI competency frameworks for teachers and students.

So during that week, we had a lot of professionals, again, from industry, from education, national, global leaders in the same room, and that was a really exciting experience to have all of these voices from all of these different angles come in a spirit of collaboration and wanting to have an open discussion, knowing that we're gonna set standards or policies or recommendations for people to reference and have a starting point.

Like, this is a pretty big deal in a way. Like, this is a unique place in history that we are in, knowing that we have this representation that we do, you know, trying to get as much equal representation as possible, like, with the resources we had. So it just was a really interesting room to be in, and it was a great experience to just hear those voices around me.

And one thing that came from those conversations was continually the educator needs to be part of the decision-making process. The educator needs to be a key decision-maker here. Like, of course, we can provide recommendations for policy at the school level or district level or state level or national level, but the educator needs to be a key decision-maker in this process.

I kept hearing that in almost every room. And it was interesting the way it was laid out because almost in a conference, you know, each room has a different topic, and some of them had panels, some of them had discussion groups. You know, it was meant to provide a variety of environments to get as much information and ideas as possible so we can Contribute to this, like, latest edition at that time, or this first iteration of the AI use policies, or rather the AI competency frameworks.

[00:40:21] Alex Sarlin: Yeah. It's really interesting to bring together different types of voices around this conversation, and I think it feeds into the concept of, you know, everybody is learning together and sort of learning in the open about thoughts about how this work-- There's no expertise fully yet. People are trying to claim it.

They're, they're moving towards it. But I think we're in a moment, even now, a few years into generative AI, where we, I think as a civilization, as a society, every country is trying to figure out their own strategy, and I think it's really important to be collaborative and to have educators, you know, at the center of that discussion.

I have a sort of zoomed out question here, but I'm really curious about your answer to it because you're somebody who obviously really revels and thinks very deeply about responsible AI use policies, about frameworks, about different types of use cases and professional learning, and I think all of it is incredibly valuable and, and we need that to implement AI, you know, effectively and correctly with impact for students and safely.

At the same time, Yeah ... something that I feel like one of the things that is most exciting about AI, and if you look at some of the people really just running with it in various directions, is the idea that it just opens up all of these possibilities. You can create new applications. You can create new stories.

You can create incredibly quickly in any different direction just based on sort of the limits of your imagination, which is a very exciting space to be. It's very empowering. And sometimes I wonder if all of the frameworks and the structure and the sort of bureaucratic checklists that are being layered onto this sort of pull away the energy, the sort of creative, empowering, inspirational energy that could be true in AI and you do see in other spaces.

I'm curious how you navigate that personally and what you've sort of seen in your work with North Star a-and within Google. It's like, how can you balance that feeling of, yes, we need structure; yes, we need guardrails; yes, we need checklists; yes, we need implementation plans and, and, and, and use policies.

But also, all of that structure can sometimes just feel like I don't wanna navigate this structure. I don't wanna have to check five different documents before I think of something. I'd rather just have an idea and run with it, which is exactly what AI empowers. How do you balance those two things? 

[00:42:30] Victoria Lansdown: Yeah.

It's a tricky one. It's a tricky one that we're seeing with a lot of players on the education side and the industry side sometimes as well. Even in organizations, there are still hesitations around the use of AI, uh, especially in financial insurance companies or on the legal side or accounting. There, there are still industries that have those questions as well.

Something we do on both sides is we start with a really small use case, and that's a pretty safe First start that we've seen. So we recognize now that we've set guardrails, we've heard a lot of these concerns, we've documented them, we have a plan moving forward, let's pick one thing that we feel good about trying out, and maybe we just start with that for a month.

Or often we work with three-month rollout plans. So let's look at over the next three months, let's test out this one use case. Let's see what worked, what didn't work. Let's come back together and reevaluate. And a lot of times, more often than not, I wanna say with a big asterisk, that helps open a lot more conversations around, "Oh, okay, this worked really well in this specific use case.

I bet it would work well in this really similar one that I can think of off the top of my head as another way to expand it. And it-- this saved me a lot of time with this one element of my planning or this, whether that's project planning or lesson planning. I bet this can also help me with some of my financial budgeting with the big asterisk next to that."

But there are a few ways that you can continue the conversation once you see something work pretty successfully and safely in a clear use case, and that kind of helps instill that creativity and that optimism and excitement, and helps that grow over time. And I think part of that too is like just being comfortable experimenting with the tools.

You know, I try to try new things with different Claude connectors or try to create new agents with Codex or I, I try to do different things. I try to find a little video that I like on YouTube as a tutorial and just try things out because-- Just find a short one or find one that you're comfortable with and then try it out.

You know, embodied practice is the best way to create those long-term myelinated pathways, is what neuroeducation calls them. And so just, you know, experimenting with the tools and seeing what the limitations are, I think that will just naturally inform future decisions, but at the same time, help encourage that optimism and creativity and excitement 

[00:44:39] Alex Sarlin: It's a great answer to, I think, a very difficult question, frankly.

One that I think is really-- I personally think it's sort of core to this moment, is AI sort of hit all of us very quickly. And as it's been progressing and evolving very quickly, more and more models, more and more open source models, more and more power in the large frontier models. It's moving into images, it's moving into podcasts, it's moving into video.

It can do incredible things, and it can feel very inspiring. And as you say, I mean, I think that I love that you're bringing the instructional design into it. The idea of actually doing it, I've noticed for a long time that the people who spend more time actually working with AI tend to be much more positive about it.

Yeah. The people who tend to be very nervous and skeptical about it often don't actually spend that much time with it. They've just sort of read the research, or they're looking at it from a structural perspective, and they're like looking at all the things that could go wrong. It's a really interesting, I think, tension right now, especially in education, which is a field that combines-- We all know that inspired, passionate teaching and relational intensity in teaching, teachers who care about their students, who care about their experiences, are trying to be engaging, have better outcomes.

So how do we balance that with the safety and the risk factors? To me, it's becoming like the question of the age right now because it's so important, and I think w- the work you're doing to really think about, okay, how do we create structures so that there is safety, and then within that safe structure, create inspiration and excitement and passion, and have that teacher say, "My passion is..."

I had a third-grade teacher who loved birds, and she did Bird Day, and everybody-- It was like this thing where everybody, whoever took her class, even here they are decades later, says- ... "I remember which bird I was using for Bird Day." And it's like that is something that AI could make even better, frankly.

Mm-hmm. But I would imagine that teacher, if she were still teaching, might say, "Well, I've done this for years. Why would I use AI if I have to go through all these checklists to make it happen? I already know it works." It's just an interesting moment. How do you react to any of that? 

[00:46:34] Victoria Lansdown: That was an incredible answer.

That was-- Yes, that is so true. I definitely can envision, you know, a similar educator with that passion, and, and that's what sticks with you over time. I-- for me, it's a fifth-grade teacher who helped, like, shape the trajectory of my life. I think, yeah, recognizing that they definitely, I could see question: Why would I use AI when I know that this works over so many years?

And at the same time, I've seen other educators with decades of experience wanna try something new and wanna explore this technology because we know that we need to prepare the next generation for this digital world. So wanting to recognize that these students need experimentation and practice with these tools in a safe way.

It's a really nuanced space to be in, I think, but it's an exciting space to be in. 

[00:47:14] Alex Sarlin: Yeah, absolutely. As you look to the future of what you're doing with North Star AI Enablement, what do you anticipate being sort of the next phases of AI incorporation, AI implementation in the classroom? Do you think that some of these policies are just gonna become sort of second nature and it's going to become a, as you say, like a toolkit and a set of best practices, and we sort of get it down and then people will lean into it?

Or do you feel like over time, as things continue to change, as new risks potentially appear, the policy layer is just gonna have to continually evolve and evolve and evolve? 

[00:47:46] Victoria Lansdown: This is a good question. I think I have two parts to my answer. So one, I'm hopeful about AI's potential to support learners and in the classroom because many of the challenges are highly individualized.

But from North Star's perspective, the immediate work and the work we can see going forward is more on the back end, like I mentioned. So helping institutions think through those privacy safe use cases and their educator workflows, and looking at approved tools, and creating staff trainings, and how to evaluate whether an intervention is actually improving their daily work.

I think that's something I'll continue to see as a big opportunity for AI in education and in the classroom moving forward is from the back end. But t-to that point, I still want to mention how exploring the ideas of play and creation are really exciting student-facing possibilities. I recently saw an example of a 10-year-old using AI to build a Lego-inspired game that educators then used in STEM classrooms.

And what interests me there is that motivation and active creation, and so not AI doing schoolwork for a student. But I'd like to see more research into whether those interactive uses help sustain attention and curiosity. 

[00:48:56] Alex Sarlin: Absolutely. Yeah, very inspiring and probably no accident that it's Lego based because Lego's whole brand is about play and experimentation and creativity, and I think that makes a lot of sense.

Really, really, really interesting. So you spent a lot of time on the ground speaking to education leaders, to organizational leaders, to educators. Um, where-- over the next year, here we are about to start the 2026 to 2027 school year. Over the next year, what do you think might change in the AI landscape? I feel like we're definitely on the sort of heading into the trough of disillusionment in many ways in schools, but there's, I think, a still a lot of excitement about the possibility.

What do you think will be different a year from now that will actually push things forward in one way or another on the conversation about the role of AI in education? 

[00:49:41] Victoria Lansdown: Yeah. I'm excited about the move from everyone using the same chatbot to people being able to build much more individualized models, agents, and experiences.

Mm-hmm. So again, we've already moved quickly from basic generative AI to agents to automation and integrations, and so the barrier to creating specialized AI experiences is dropping quickly. And in education, that creates room for something much more interesting than like an AI-powered general purpose tutor.

We can now create specialized models- in agents around a district's policy or an institution's knowledge base or a department's workflows or a particular professional role or an Edtech company's product and customer support ecosystem. And that opens up much more interesting possibilities than just giving everyone access to a standard tool.

So I think the next era or next year of Edtech will be less focused on these little demos and more interested in the evidence by looking at data like what changed after I implemented AI? What does this cost to maintain? Am I starting to look longer term? Where was human review still required so I can continue to operate this safely?

And did this create measured value compared to the simpler process it replaced? 

[00:51:01] Alex Sarlin: Very, very interesting. I love that direction. I think that idea of one thing working backwards from impact and value and actually not just playing, but actually doing things that have a very specific goal is, is core. But I also love what you're saying about that idea of moving away from everybody using the same tools and trying to custom fit it to many different use cases to having more purpose-fit tools and more purpose-fit workflows and agents in all sorts of ways.

I feel like it's already begun, but I think w- there's a lot of room to go, and I think the Edtech industry has a lot of really interesting roles to play in that exact set of changes because as people move away from just using one off-the-shelf model for everything, they might start to say, "Okay, well, what actually is gonna get us to solving the problems?"

And, and there may be some custom-built solutions for exactly that. I appreciate this conversation. It's really, really interesting. So last question for you, what other resources would you share for our audience that you feel like would be useful for them in navigating this moment in education? 

[00:51:59] Victoria Lansdown: Well, I like to listen to "The Learning Experience Op Show," which is a podcast series hosted by Jack Rabbit, where they have conversations with professionals who build and run learning systems across sectors like corporate L&D or higher education, K-12, and healthcare.

You know, similar to Edtech Insiders, which is another resource I recommend. But that's been helpful for me to k- help keep my finger on the pulse in this space. And then a few I mentioned already, but just wanna restate is, one is neural education. So if someone is interested in the learning science side, especially attention, cognition, regulation, and thinking more carefully about what technology is doing to the learning environment, definitely would recommend looking into their research.

And then Ednology or Ednology experts if you're interested in the implementation and ecosystem side, how schools find products or expertise or partners or funding. I am very excited about how Ednology offers funding as an avenue to break down that barrier, and they provide, you know, practical support rather than just looking at Edtech products in isolation.

I just also would wanna say like to continue to follow your local school district and state education news. Uh, some of the most important AI changes aren't coming from OpenAI or Google. They're happening through district guidance and procurement decisions and state legislation and classroom policy, and what educators are actually being allowed or asked to do.

I would say continue to ask those kinds of questions and look into those resources. 

[00:53:21] Alex Sarlin: Fantastic. Neural education, Ednology, "The Learning Experience Op Podcast," is that what it's called? 

[00:53:26] Victoria Lansdown: Op Show. Yeah. "Learning Experience Op Show." Really good podcast. 

[00:53:30] Alex Sarlin: Fantastic. We really appreciate those recommendations. And as always, we will put them in the show notes for the episode.

Thank you so much, Victoria Lansdown, MBA, founder of North Star AI Enablement and contributor to AI for Educator programs at Google, Pearson, UNESCO. Really appreciate you being here with us at Edtech Insiders. 

[00:53:49] Victoria Lansdown: Thank you so much, Alex. It was great to be here. 

[00:53:51] Alex Sarlin: Thanks for listening to this episode of Edtech Insiders.

If you like the podcast, remember to rate it and share it with others in the Edtech community. For those who want even more Edtech Insider, subscribe to the free Edtech Insiders newsletter on Substack

This season of Edtech Insiders is brought to you by Starbridge. Every year, K-12 districts and higher ed institutions spend over half a trillion dollars, but most sales teams miss the signals. Starbridge tracks early signs like board minutes, budget drafts, and strategic plans, and then helps you turn them into personalized outreach fast.

Win the deal before it hits the RFP stage. That's how top Edtech teams stay ahead.