How To Embed AI Literacy In Your School: Practical Steps To Get You Started [Webinar Replay]

Good afternoon everyone and thank you so much for joining us today. I just did a quick introduction to myself. I'm Rob Langman, I'm the Chief Commercial Officer here at Third Space Learning. Really pleased to welcome you to what I think is going to be a genuinely valuable conversation focused on AI literacy. Before I introduce our guests, let me just set the scene for why we're having this discussion right now. So just a week before last, the government published its response to Professor Becky Francis independent curriculum review. For the first time, AI literacy will be mandatory in English schools from 2028. Embedded across primary, secondary and post 16 schools now have a clear policy mandate to build AI literacy among staff and students. However, while the policy framework arrives in 2028, AI is already being used in schools today with some, you know, there are stats out there which show up to 85% of students already using AI. Teachers are experimenting with tools like ChatGPT Gemini for lesson planning, administration tasks. Students are using LLMs for homework, coursework and school leaders are being asked to make decisions about what's acceptable and what's not, often without much support or clear guidance to work from. We know this because schools are coming to us for advice. So at Thirdspace Learning, we've been working in the space for longer than many. Lots of you may already know us. We're not a tech company that's jumped into education. We've always been an education company which utilizes tech to help help students improve their learning. We delivered 2 million tutoring sessions to pupils across 4 1/2-2-000 schools over the last 10 years. Many of you know, 99% of schools have used some of our free resources and content. We've now used that experience in developing high impact online tutoring to develop sky, our spoken AI mass tutor. We're doing it the same way teachers are leading it, designing the lessons, the pedagogy, but with human oversight or with, we may hear us refer to it as human. In the loop on sky from day one, that's obviously taught us a huge amount about what it actually takes to use AI well in schools. Not just the technology, but the leadership, the culture, the ethics, the safeguarding and the classroom practice. And we realize that whilst there's a lot of noise out there about AI, there's not enough for the people who actually need to make decisions about, to make decisions, school leaders. So for those school leaders who are looking for values led practical guidance, we've created something to help and which Laura has led. Today we're launching our AI literacy course for school leaders which is completely free. It's designed specifically to give heads, deputy heads, mat leaders, and to be honest, anyone who's interested in learning more about AI literacy and how it's working in schools and. Or anyone who wants to be an AI champion in their school frameworks to lead AI adoption responsibly, strategically and in line with their school's values. Talk us through that work and have a really honest. I'm calling it a Fireside Conversation. Fireside chat, which, given the weather, feels appropriate and given that you've got a spiced Ribena drink, feels very wintry.

Absolutely.

I'm delighted to introduce Laura Knight. So Laura is an education consultant. She's a former senior leader, Digital Education and AI Specialist, Tech Women 100 Award Winner, 2025, who has spent her career thinking deeply about curriculum, pedagogy, school, school improvement. She's worked with schools and maps across the country. She's been at the forefront of thinking about how AI can strengthen teaching and learning whilst respecting the professional craft of educators. Laura led the design of this course and I think you'll find her thinking is both pragmatic and principled, which is exactly what we need right now. Laura, welcome.

Thank you.

Okay, so thanks everyone for joining us. We do have the Q and A open. We've got a load of questions which have been pre prepared from schools who've sent us them, so we're going to tackle those first. If anyone wants to ask any questions, just pop them in there. I'll try to get around to them. And we might be tight for time. We're going to try and run to 30 minutes or so and if we don't get around to it, we'll have a read for all of them and we may use them for a Future webinar, blog, etc. So don't feel free to pop them in there. So let's, let's. Laura, let's open up with the moment that we're in. The government's just responded to Becky Francis curriculum review. AI literacy is now mandatory from 2028. But for a school leader who's watching this, judging 100 priorities, why should this be on their radar now rather than in three years time?

So it's a really interesting question and I think there are lots of school leaders who are going, yeah, I can see why that's important, but I've got a million other things on my plate, so this is going to have to wait. Now, I would argue that it's worth pushing back against that and thinking instead about what you can do now and why that's important. The why it's important, I think, is key. First, I think there's three key things here for me. The first is what I would call drift, the kind of lack of focus, innovation that will happen very organically and is probably happening anyway in your school, where folks are using tools without structure, without policy, without guidance. They're doing so in a way that's patchy and inconsistent and that's unhelpful from a compliance perspective and from a governance and accountability perspective. But it's also just rubbish for kids who, I think pupils and students deserve something more consistent and better structured. The second piece is around duplication, and that inconsistency that I was talking about can also then appear in classrooms. If you've got one department doing one thing, another department doing something else, what they're told in year seven is then completely cancelled out in year eight. That's really unhelpful as well. And the final thing is around dependency. I worry about habits and how important it is to set powerful intentions when it comes to good practice with technology. And this isn't just about AI, this is about all of it, but particularly when we're thinking about AI, there are a whole bunch of things that I think are very tempting for teachers to explore entirely, understandably, but we do have to be careful about boundaries and guardrails and make sure that we're not sending people down into pathways where they're trying things that they don't fully understand and they don't understand the implications. They don't understand what might happen as a result. And so it makes sense for me to think instead a bit more proactively and see what we can put in place sooner rather than later.

Yeah, yeah, definitely. And like. And obviously they're all using it. They are using it now, so we can't wait for a year time.

Absolutely, absolutely.

So when you spend a lot of your time talking to school leaders about AI, you're always speaking at most education conferences. What's the most common thing you hear when you talk to school leaders? Is it excitement? Is it anxiety, confusion or something else entirely?

So often it's about balance. And there is, I think, a real concern that there's a lot of division and a lot of, I say, polarized thinking in this space. And headteachers are having to sort of fight their way through and work out where they need to sit on that spectrum, where you've got doom and gloom and disaster and, you know, block it, ban it, throw away the key, kind of at one end, and then you've got evangelistic, almost sort of blindly naive techno joy at the other, and they're having to balance a whole series of competing demands and agendas when there's very little clarity about what excellent looks like. Now, if you add into the mix that a lot of leaders entirely understand, understandably, don't necessarily have a great deal of confidence themselves in this space, they wouldn't necessarily consider themselves expert. They haven't got huge amounts of time to sort of sit and experiment and play themselves. So it can be very tricky for leaders to actually have give themselves some grace around that sort of beginner's mind piece. But also, I think it's so important that we consider how leaders need space and time to grow and learn themselves, rather than just hoping that the head of computer science will sort it out or the director of IT or somebody else's problem. What I think is important is that this piece around AI and digital strategy is a leadership issue. There's the technology aspect to it, but it's a leadership issue because it touches so many aspects of school Life.

Yeah, yeah, 100%. And you touch on the idea that it might be tempting for some leaders to say, maybe this is one for the IT department or the computing teachers to own. We've designed this course specifically for school leaders. Why then? Why now? And I suppose once leaders have this literacy, how do they build it across their staff, their students, the wider community?

So I feel very strongly that in order to make the very best use of this technology, it needs to be aligned with the purpose and values of the school. And it's absolutely the case that it should be the leadership of the team, the leadership of the school, who are curating that whole experience, that culture, that sense of ambition about their community. So, first and foremost, as I say, it's a leadership issue and it should be aligned with the purpose and values of the school and the context of the school. Let's not imagine for a moment that a teeny tiny rural primary is going to have the same approach to this as a huge urban secondary. Like, they're going to need to be contextual approaches to this that are very specific. Specific. And then in terms of how you move forwards from there, I think giving to, giving leaders an opportunity to learn, to explore, to question. And I think the questioning part is key here. We're not saying in this course that we're going to give everybody all the answers. That is certainly not our goal, wouldn't be so arrogant. What we would like to say is that we're encouraging people to have great questions and to go away from this course feeling confident about where they might go next. There is so much practical stuff in there about things you can try, what you can bring up in your next meeting, who to go and talk to about this next step so that leaders feel confident about not only the principles that we're suggesting, but also the practicalities. And I hope that that's something that's really valuable.

Yeah.

And I think.

I think the fact there's also, like, downloadable resources which comes alongside the course for each module, which certain toolkits, etc. They're really useful as a starting point because they give you. Give you a launch pad from where you go, okay, let's use this in our next.

Absolutely. And I think it's really helpful as well for colleagues to have something extreme external, because we often find that people can have quite an emotional response to innovation. They might feel insecure, they might feel protective, they might feel, no, AI is my thing. I'm the one who does this in the team. So to have something that's produced by a highly regarded external party, in this case, and to be able to bring it to school in the sense that it's designed to be helpful, it's designed to be adaptable, and then the schools are not starting from a blank page or from a highly personal view. It's instead something that can be really adapted for their needs.

Yeah, wonderful. And let's talk about. Dive into the actual modules here. So module one of the course. You talk about six fundamental shifts that school leaders need to make. One of those is moving from defensiveness to confident professionalism. Could you just unpack that a little bit for me? What does defensiveness look like in schools right now?

Sure. So I think this is a really interesting one. Headteachers and senior teams in schools and in trusts have to be sort of Cerberus at the gate in so many ways. They have to have a really protective stance about. About their school, their community, their routines, their staff. Because in a situation where resources, budget, all of these things are limited. You have to protect what you've got and really be careful about the decisions that you make. We don't typically kind of go fast and loose with decision making and innovation in education. We don't have the resource, we don't have the time. And we also care deeply about our students and want to make sure that what we're choosing for them is the best possible choice that we can make with what we have at hand. And so that defensiveness is I am under no circumstances am I attacking that it is a totally normal and responsible choice. In fact, I would actually say it's a more responsible choice than the kind of gung ho, let's go for it and damn the consequences kind of approach. So defensiveness is fine as long as we can see beyond it. It is a phase rather than a permanent state. And actually the way that we move through it is through learning and understanding. And that isn't about saying, okay, you've got to sit there and learn to code, you've got to build your own stuff, you've to be super technical. That's not at all what I'm saying. Actually. Quite a lot of this is about philosophy really. It's about leadership, it's about understanding culture and people. It's about thinking about change and disruption and it's a thinking about how you apply something new in terms of innovation to your context and make it work for you. So that moving from defensiveness is about that journey that leaders are going to go on having been informed on some of those key ideas and some practical approaches to help them bring their team into a conversation which is really forwards facing.

Yeah, 100%. I think experimentation and modeling, getting staff members who have successfully used AI to model for others, that applies to more than just the school context is business.

It really does. And you know, some of the things that I love and some of the things that I've learned from colleagues past is, is the, the absolute power of somebody who would not consider themselves the wizziest, flashiest expert at something, but having the courage to stand up and show their practice with humility and grace like, and showing what they, what they've done. Because there is something so powerful about seeing it in action and understanding someone's process. And I think if we can see that process rather than just here's what you get, this is why how I thought about it, this is and that, because then folks can model, use that model to help, you know, adapt it to their own situation. And it sp new new thoughts. But also, and I think this is where I was talking a little about that vulnerability piece earlier. It shows such leadership strength when a leader is prepared to fail forwards in front of and with and collectively, you know, with their staff. So if you can do something, try something out, maybe it doesn't work perfectly first time, but you model how you move forwards through that and say okay, that, that wasn't quite ideal, let's see what we do next. Rather than jazz hands drama, you are modelling not only the new tools and the new approaches. But you're also showing your own capacity to iterate and showing the behaviors and attitudes you're expecting as well. And I think it's hugely powerful to see that from whether it's from the stage, from the staff meeting kind of chair, wherever it happens to be.

I think, yeah, open vulnerability and the ability to say, I don't know is fine. And that should always be accepted in leaders.

Yeah, yeah. And giving ourselves room to maybe sometimes be playful about that, but knowing where the edge of play and experimentation should be. One of the things that I worry about, and I will just mention this now, is that I think that stage one of our exploration of AI should have plenty of that kind of exploratory, experimental piece in it. But we can't stay in that phase forever. It's scrappy, it's messy, it's organic, it doesn't have roots, it doesn't have intention. So let's make sure that we have purpose behind it, which is much more anchored in what we really want to achieve.

And you talk of purpose, so purpose, not policy, is a theme in module 2. What's the difference between a good AI vision and a weak one?

So I often talk about the tail wagging the dog here, and there's a little bit of kind of keeping up with the Joneses in this space where, you know, the school down the road have done some fancy thing and therefore we should do a fancy thing. Can you do something fancy marketing team like? And actually, I don't think that's that helpful. So we don't want the tail wagging the dog. We don't want to do things for appearances sake. We don't want to also create busy work for staff that doesn't have real impact or any meaning. So for me here, this is about looking at your context and asking some very specific questions. Once you have an understanding, understanding of the technology, something like, well, actually, when we're looking at this, what are the problems that we want to solve? What are our number one issue? When we look at our strategic plan, we look at our development plan, like, what do we need to be doing here next? Let's think about what opportunities we can now create with this technology that we couldn't before. What does it unlock? What does it enable? What does it shift? Maybe what can we let go of? Because, goodness, we don't have to keep just adding more. And then also, when we're thinking about those opportunities that we want to create and the problems that we want to solve, there can also be Something creative there. Something that is a little bit more about. Well, actually, is there an opportunity to think about this as a way of injecting more joy into school life? I know it sounds a little idealistic, but, you know, indulge me here in a space where. And the latest report, you know, in terms of teacher satisfaction and well, being disastrous is very, very bleak. Very bleak. There are so many folks who are stressed, overwhelmed. Their mental health is, you know, through the floor. If we can use this technology to help staff feel that they are able to access new, I don't know, new productivity, of course, but actually new creativity, new opportunities for expression, new opportunities for engagement with their students, that unlocks some professional joy. I think that should be way up the list of priorities. I mean, efficiency is nice, right? We like efficiency, but it's not the only thing. And nobody is counting hours when they're working in the way that some tech companies might imagine that they are. Teachers never have enough time. So nobody's going to be sitting, putting their feet up on a Friday afternoon, doing the crossword because they've run out of work to do. That does not happen.

I was going to say, I think there's a. If you look at the results the English education system gets for the resources it's given is pretty remarkable. So they're pretty efficient. Yeah. Let's talk about resistance and separate and after that, parents. So not everyone's excited about AI. Some, some people are obviously worried about professional identity, future of teaching. How should leaders respond to that resistance without either dismissing it or getting stuck in it?

Okay, this is, this is super interesting. The evangelists and the keen beans will find it really annoying that there are folks who are resistant and will see them as laggards. And I think that's unhelpful because professionally it is our responsibility, like I mentioned about the Cerberus at the gate thing, to maintain and safeguard the, the very best of our profession and the work that we do, as well as the well being and safety of our students. So for me, the key for me here is how do we translate fear into responsibility? How do we move beyond just a blanket kind of emotional response into something that feels much more rational, where we can look at the pros and cons, we can think about balance, we can approach it in a way that is measured rather than just anchored in that kind of blanket, oh gosh, I don't know what we're doing. So we're just going to not. Or we'll do everything and hope for the best. I mean, that's another option. Not sure that's a great one either. So yeah, that piece about responsibility matters hugely. But also I think this, there's a, there's a section of work in here that's about understanding the truth of what's happening in your context. There's a lot of what we would call shadow it kind of practice where staff are not comfortable sharing what they're doing with AI with their peers. Last month in October, teacher TAP did a great survey on this and it was A. About 57% of staff questioned said that they would be comfortable sharing what they're using AI for in school with their colleagues. Yeah, that's a big wedge of folks who are like, nope. Yeah. Now for me that says a lot more about culture than anything else. And this, this, there's a whole bunch of issues around integrity, around suspicion, around what, you know, professional responsibility and reputation wanting to be seen to do the right thing. So there's, I think again we come back to modeling but also policy on what it means to do a job where the boundaries are, you know, what are we going to do when the adults do the wrong thing? With AI in our schools, we're probably quite clear on what we should do when the kids do the wrong thing. Like that's fairly straightforward. We know about cheating and stuff. When the adults do the wrong thing. Is that, is that code of conduct stuff? Is it. It aup. Like how, where does this live? But also there is a kind of old fashioned macho, I want to say almost like martyr like thing about suffering and education. If you're not, if you're not suffering, if you're not almost making yourself ill, then are you doing it properly? And that has no place in 2025 and certainly not in 2026 when recruitment and retention is where it is and staff well being is where it is. So what we should be doing instead is looking at where we can move beyond that fear into responsibility and also productive collaboration. So we should be sharing prompt live groups with colleagues and saying, I found a really great way to do this. Here we go folks, let's try this one out. Can anyone help me improve it? That for me feels really powerful and then having absolute clarity. The edges are so for example, pieces that are non negotiable human should stay and be very. Everyone should be clear that for example, we never use a large language model to write an email to send our condolences. That's gross. Like that for me feels like if it's something that needs to be said, let's Say it. But let's also be clear. If we're sending transactional emails that are really just about making sure a message gets sent, do we need to be signing that off by hand? Like let's think it through.

Yeah, yeah. To completely agree. You mentioned a lot of the work school leaders ask you often about how to handle questions that parents may have around AI use in school. What are the messages that matter the most?

This is super interesting and again this is so contextual and depends on your parent body. It is impossible, as I have learned in many years of working on digital strategy, one to one devices, all of these kinds of things. With school schools it is impossible to make everybody happy because you will have all the colours of the rainbow in terms of enthusiasm versus reticence for some if it'll be not enough and for others it'll be too much. You can't win all of it. However, where you can win is showing that you are making considered decisions based on fundamental caring about the child and their learning and their progress and their safety. That comes from things like clear rationale, shared language, a really comprehensive FAQ actually that says we thought about all of these things and we care enough to make sure that they're written down for you so that when you ask we've got somewhere to direct you to. So part of its communication, part of it is also then really being clear on where the values are and how they translate for families. So what we understand about what it means in the classroom might be very different for folks who are are brand new to secondary school, for example, and don't have any idea about what this means. The pieces I reckon are going to be challenging are around age appropriate autonomy and consent. This is still murky, this space and there are lots of schools who are kind of trying to work out, well, okay, well we're currently a Microsoft school and we want to use this and we could copilot, but maybe also is Gemini better? Like there's all these questions around, well, what are the right tools to put in front of children? What are the right tools to put in front of staff? How do we keep every our data safe? It's complex. Parents want to know that we're taking a, you know, a caring kind of in line with school values approach. Yeah, the piece around consent is hard because the moment you ask for it is the moment that we need to have an answer for. When people say no. And I think we should be respecting agency, personhood, all of those things, but where you have solutions that are genuinely difficult to get your head around as a teacher or a leader. It's not reasonable to be asking parents to do that work for you because it won't work well. So, yeah, lots of empathy, I think, for families and. And just know that whatever you're doing will end up in the WhatsApp groups regardless, so you might as well make it good.

What. What we've seen work and I've actually been amazed by how, you know, little resistance we've had from parents in terms of them. When schools have been integrating SKY and AI tutoring to them has been. We give schools a letter to provide to the parents, which. Which firstly, it frames it in a positive. This is an opportunity and an arc for them that they otherwise would not have access to. And so for a lot of parents, it's. They can't afford to. To pay for.

Absolutely. And so you're making something available that they're just. They just couldn't have access to in another way. And that. And. And that's a very compelling starting point. Right?

Yeah, yeah. And then reassuring them on data protection, safeguarding and just like all the key questions to show that the school has thought about this and had this discussion.

Yeah, yeah, absolutely. And I think one of the other things to think about is the contextual fears that people have. Remember, a lot of people get their news in teeny, tiny bites. They will not read the newsletter, they will not read the briefing, they will not come to the webinar, they'll have the. The comments that somebody made in the WhatsApp group or the forum. And so we have to be ready to deal with that fragmentation in a way that's powerful and timely. And I think, particularly when parents are just regurgitating tiny sound bites, like, I get a lot of kind of water boucherie, which, it sounds a bit dismissive, but, you know, we get a lot of. Well, what about the water? Because Sam Altman said a thing about water in an interview at one point and it's really unhelpful because in terms of water consumption, like, beef is 100 million times worse, golf is exponentially worse than AI ever will be. And this is. But we never think about these things in context, so we have to deal with that fragmentation with care and make sure we're really well informed.

Yeah, 100%. And you mentioned risks, what keeps leaders up at night? Let's talk a little bit about, you know, safeguarding, bias, hallucination, the common. The common three. So, yeah, module four tackles the risks and limitations head on when it comes to safeguarding data. Protection, inappropriate content. What's the one thing schools are most likely to miss or underestimate?

I think it's synthetic intimacy, which is one of these things that's around relational aspects with AI and how easy it is for youngsters to end up in an almost dependent, I'm going to say relationship in inverted commas with an AI like a chatbot of sorts, and that those things are sort of sold to them as confidants, companions, virtual girlfriends and boyfriends, that kind of thing. And I think the problematic nature of this and this sexualizing of some of that content in ways that's really inappropriate and how, how poor even now post Online Safety act, how poor the age verification systems are in so many of those spaces is a real worry. So I would say as a heading, synthetic intimacy is a big deal. The other thing that's on my radar is naivety around data. And I think this is something that we talk about in the course and that question of if you're using a free tool that is, you know, you've just signed up with, essentially you're paying with your data. Right. And so, and there's a naivety around where that information goes. And so I would like for, in terms of literacy, to really focus in on helping youngsters understand how important it is to protect their privacy. By all means, explore free tools like Go Go for it. But remember that whatever you're doing with that, your eyeballs and thumbs are for sale in some way or other in your inner thoughts, your needs, your desires, your fears, like all of these things that you're expressing. So I would like to think about self awareness and self regulation in that context.

Yeah, 100%. And exactly what you're uploading. If you have, like you said earlier, if you've got individuals all over using different tools, they're all using the free version. Everything they upload is probably, is probably likely going to those companies to train their models. If they're paying for it, then they can't use it to train.

Absolutely. And there is a challenge here because we will never see AI large language models improve to meet the needs of our highly contextual individual needs if we never give them that data. But it needs to be done ethically and it needs to be done with the right consent. And I'm afraid when we think about the history of things in terms of social media and the sort of surveillance capitalism approach to this stuff, it's, it's a little bit tricky to have trust. And unfortunately there is this sort of entirely understandable apathy from younger people about the meaningless nature of all of that legalese and terms and conditions and, you know, all of that.

It's a great point. Bias. So you talk about equity bias, inclusion, essential sort of central concerns. Can you give a tangible example of how AI bias might, might show up in schools? Not in like a dramatic way, but in the everyday decisions schools are making.

So let's say, let's say, for example, we're using an AI tool for resource creation in teaching and learning. We're planning some stuff for our students. It is likely that there, without prompting for diversity, for example, you will end up with a very specifically narrow range of characteristics that arrive in your case studies, your examples, your imagery in particular. It is still the case now that if you ask generative AI to produce imagery for you, it will be almost poetically stereotypical. It's almost like a caricature. In some cases it's improving, but it's slow. And so one of the things that we can't expect by default is that those approaches to bias, particularly geographical and linguistic, you know, all of these things are. We should be thinking about richness and diversity and especially those who are in international contexts. If you're. I mean, I've just come back from some work in Ghana, for example. If you're in West Africa, you don't really want examples from Welling Garden City, like, oh, California. That's not relevant to you. So we do need to think about localization. We do need to think about our context and making sure that our young people are represented in the work that they, they see. And so that those examples, those case studies, all of that is appropriate for them. And it takes just a little bit of effort to make that happen, but we shouldn't expect it by default.

Yeah, I think that's a great example. Let's move on. Hallucinations. Now, AI tools. Everyone's noticed mistakes at points when they've been using them. AI tools can generate plausible sounding content that's completely wrong. Wrong. And what is often referred to as hallucinations. How should leaders build safeguards against that without creating so much friction that staff stop using AI altogether?

So I think we have to be pragmatic about this. First of all, there's a professional responsibility when it comes to quality for us all to make sure that we're not just parking nonsense in front of kids and hoping for the best. In the same way that not so very many years ago when YouTube was new, you wouldn't just throw a video on without watching it first. This is kind of teaching 101. And in the same way, we shouldn't just be copying and pasting blindly without checking and without making sure that the content we're sharing is accurate and all the rest of it. Now, this is hard because there are many large language models that take an approach where you don't have great transparency on the sources and where the way things are framed sounds so confident and so slick that it's really easy to be sort of seduced into a sense of. Of trust. And so we do need to be realistic that there are going to be things that come up that aren't real. I suspect the moment that we're looking for particular examples for, you know, resources that actually rely on data, any of that kind of thing. This is important. Just today I was working with some colleagues on aggregating quotations from a piece of work that I sent a set of questions to eight different people. They are sending their quotations back for those questions. It has mixed up who said what, and if I didn't know, if I hadn't actually read it, I wouldn't know that. John didn't say that, Sarah did, because it sounds so plausible. So we have to remember that mimicry is a great skill and accuracy is not. And there's lots of talk about reducing. Reducing hallucinations. And the reality is that the foundational technology we're dealing with, with large language models alone, is a predictive tool. It's a big statistical engine that just knows what's likely to come next. And as it sounds, sounds close enough, that's fine. As is the case with sky, if you want to make sure that something is accurate, it has to have something more foundational that is going to anchor it to reliable content and to give it a context to operate within that is solid. And that foundational piece is absolutely key. And that's where we should be looking for quality. So where accuracy matters, check it where it doesn't. Be careful still. But, yeah, let's not make assumptions that are unjustified. Yeah.

The only thing I'd add to that is that as humans, we. We get things wrong all the time, and we don't. And reduction's important, but it's the point you made earlier. As long as we're not playing knowingly inaccurate content in front of students.

And this is the thing. Yeah. So I think it's about how high the stakes, essentially. If the stakes are super high, don't be a Muppet about it. You can quote me on that. But, like, I think it matters if the Stakes are high. We should be taking professional responsibility for checking accuracy. If not, then perhaps we can let it slide. And like you say, humans make mistakes, but they're not the same sorts of mistakes. We're typically pretty good on context, unlike AI. And so that's where the disparity sort of shows. But it shows especially, and this is one of the fundamental issues here, if there isn't good transparency and sharing about the processes you've gone through and the AI that you've used. So if you're one of that big percentage who can't talk to their colleagues about what you're doing, it's going to be very difficult to climb down from that. I wrote this myself and actually no AI has hallucinated it all and I didn't check. That's where the real sort of dissonance is.

Yeah, wonderful, wonderful. Let's make this practical. So school leaders are being pitched AI tools constantly and they're constantly experimenting with them. What's the one question a school leader should ask every vendor that will cut through the marketing and reveal whether a tool is actually fit for purpose?

So I, well, there's a, there's a bunch of these questions in the course. And so I think it's worth going and having a look at this section because I think it's important. I feel very strongly about the data and about protecting data and protecting the autonomy, the privacy of the young people, because we have to essentially in this situation, consent on their behalf when we're looking from a procurement perspective. So I think for the most part, let's begin there. Like, is the data stored responsibly? Where is it? Where did the training data come from? How appropriate is this? You know, those, those kinds of really nuts and bolts questions are a very good place to start. We do then have to think about pedagogical value because unfortunately there are lots of tools where the jazz hands and the shiny coolness of it might be really impressive. But is it actually going to deliver value in the classroom or is it going to deliver value outside of the classroom in a sort of, you know, after school capacity or whatever it happens to be? We need to see evidence of that. Now, I'm pretty tough on this and I want to see case studies, I want to see impact, you know, research. I want to see like, where's the, show me the actual, like information you got about the, the real work that happened here. I'm quite demanding with companies about that. That's tricky for a startup. But then would, would you trust a brand new Startup to roll it out across your school. You shouldn't. No, they should. They should have like you. You should be able to work through that in a way that feels very measured. And so trust needs to be earned in this space.

100 and there's lots of places where they can access the evidence and the research. CEF is a great place to start.

Absolutely. EF is a great place to start and there's lots of great networks as well with colleagues sharing all sorts of good stuff here. So we should be making sure that we can sort of listen to colleagues and not work in isolation.

Wonderful. Let's talk about the AI literacy course that launches today. And I should mention, it's completely free for school leaders. Walk us through what people will actually experience when they take this course. What does the journey look like? What will they walk away with? How can leaders access it?

Okay, so this is. It's not going to take you hugely long. It's snappy and it's about an hour, isn't it, Rob? I think. And so it's done in a series of chapters and segments, if you like, that include lots of inputs, some provocations, some good questions and then lots of actionable suggestions about next steps that you can make. There's downloads and documentation that go with it so that you can take those materials back to school and have conversations with peers, with your senior leadership team, even with governors. And we need to make sure then that we are sharing it generously because it's free. And so please feel free to make sure that colleagues are accessing it and can get the benefit there, because I think that's a really great opportunity.

Wonderful. Let's end on a forward looking note. When AI is done well in schools, the right technology, the right safeguards, the right pedagogy, what becomes possible that wasn't possible before?

So I think this is, for me about agency autonomy and lifting the lid on limitations. That's what it's about. It's about giving every student new opportunities to. To access learning, but also to kind of spread their wings in a way, whether it's through creativity, through planning, through send support, through collaboration. There are so many tools that AI can help with there, and I hope that that's a really sort of exciting opportunity for youngsters who I think should feel like the only limitation on them is their curiosity here.

Yeah, yeah, yeah. Wonderful. For someone watching this, who's thinking, I need to get ahead of this, what's the single most valuable thing they could do this week to start building AI.

Literacy in Their school come and do the course, surely. But yeah, come, come and check out the website. Come and have a look at this, see what you think. Like I would just, just go watch the first video and see what you think. I think that's possibly the, the, the power move here, but also just committing a little bit of time. If you said I've got 10 minutes, I'll go make myself a coffee and come and have a look at this. It doesn't have to be a hu. But starting rather than putting it off is an important thing. So yeah, I think that's probably the best place to begin.

Yeah, definitely. Let me just walk through what's included in the course. So it's completely free. The web link is here and what we're going to do is we'll send you everyone a recording of this. We'll also email this web link to you so you can access. Modules 1 and 2 are available right now. Module 16 AI shifts for school leaders. Then it's leadership, vision and culture. Modules 3, 4 and 5 are coming in the next couple of weeks. We released. It's designed for school leaders but it's not limited to school leaders. Share it widely. For anyone who really wants to learn more about, you know, AI use in schools, anyone who wants to be an AI champion, I think it's a perfect place for them to start. This is what it looks like on the website. You can access it all for free. It's five modules. It'll probably take an hour in total to go for it. You don't have to do it all at once. It'll save your progress. Come back whenever you have 10 minutes. You also with every module you'll see some downloadables which are free of charge. These are things like the Rose and Shine checklist of lesson evaluation questions, vendor questions to assess safeguarding compliance, data controls for new tools, decision frameworks for any AI adoption in your school of trust staff development scenarios. You get a certificate of completion for each module. Feel free to share that and post it with the course along on social media, LinkedIn, etc. And the first two modules are live now, will be available as soon as possible. The only thing I would add is that if, if anyone wants to also see what an AI a student facing AI tool looks like in school. We always offer free sessions for schools, for sky and it's a good place to start. Do a few free sessions, see what it's like, get student feedback. You'll learn a lot about how students interact with AI doing that. No obligation to do anything afterwards. We're going to wrap up there, and I thank Laura very much for attending, and we will speak to you all soon.

Yeah, wonderful. Thank you ever so much.