Cracking the Code: How 8 Simple Steps Unlock AI-Driven Change in Schools
If you're worried about how to lead this massive change process into next year as AI is set to take over the world, two leadership gurus have got your back.
If you’re a leader right now you’d be forgiven for thinking ‘where do I even start?’ when it comes to introducing and embedding AI into your school or business. The problem you have is that it’s so new that nobody really knows the best way forward, but it’s accelerating so fast you don’t really have the option to ignore it.
Thankfully, leading AI change isn’t much different from leading any change process, and we’ve plenty of experts who've written on it over the years. I’m going to explore how two of the best writers on this subject - John Kotter and Kurt Lewin - will be a huge help to you as you navigate the coming year. Kotter in particular has been enormously influential on my leadership of turnarounds over the years.
Eight steps to leading change
In Leading Change, John Kotter gives us almost everything we need to move our school or business forward into the future. He outlines eight steps that an organisation needs to go through, stating that it’s important to move through them all one by one: failing to complete one before moving onto the others can undermine the entire process.
1. Establish a sense of urgency
Think of any high stakes thriller. They are most impactful when the hero is in absolute jeopardy; when they are backed into a corner and seem to have no way out. It’s the same when managing change. You can only effectively drive through your plans if you make people understand that there is no choice but to change what they’re doing. Establishing a sense of urgency is therefore the critical first step. If your team believe that your plans could just as easily be actioned next year, they won’t see the need to rouse themselves to action.
The shift away from the current educational paradigm is coming, so I don’t think this will be too difficult to emphasise. However, teaching is no different from other sectors: there is an established way of doing things and most would prefer not to rock the boat. Making your staff understand that there actually isn’t any choice is your first task. How do you do that? By impressing on staff how quickly the landscape is changing.
Having statistics to hand is advisable: the fact that ChatGPT scaled to 100 million users faster than any technology in history, the incredible speed at which AI is improving, the number of students who are using the technology already. You may wish to begin by anonymously surveying students so that staff can see how many of them are using it in place of their own work. Whatever approach you take, what staff need to see is that they don’t have a great deal of choice: AI is here to stay and is only going to become more ubiquitous and unnoticeable, so better to embrace it and learn all there is to know than bury heads in the sand and pretend they can carry on as before.
Kurt Lewin is a useful adjunct to this. In stage one of his Unfreeze-Change-Refreeze model, we focus on creating awareness of the current state of education, how ill-fitting it is for both the present and future, and how AI can support making much-needed changes. It’s important not to make AI the sole driver of change: remember how Mark Zuckerberg pinned his entire company’s future on the metaverse, even changing its name from Facebook to Meta? Rather than looking more broadly at how the worlds of work, education and leisure might change in future, and how the metaverse might support this change, Zuckerberg focused his lens entirely on one technology. More than $30 Billion later with little to show other than a couple of hundred thousand users of his clunky virtual world, Meta is now focusing more of its attention on AI.
It is far better to raise awareness of the skills children need heading into the future, and how technologies like AI can best support. A good place to start is the World Economic Forum’s Defining Education 4.0 paper, released in January 2023. This outlines the skills children will need in order to be successful in the 21st Century and will therefore give you an indication as to what concrete steps you will need to take in order to ensure every skill is covered. AI will help across the board for sure, but it is not the answer to every problem nor should it be. If there is one thing I hope this book gives you, is that AI can and will be a huge driver of change, but will not herald in a new dawn without our collected efforts to use it wisely and to ensure we foreground our essentially humanity at all times.
2. Create a guiding coalition
The first thing you need to realise is that you won’t be able to do any of this alone. Shouting from the rooftops about the necessity of change is useless unless you have a team who can help you drive change though. You need your people around you if you are to have a hope of being successful. No matter at what level you are in your organisation, it is critical to ensure you have decision-makers and influencers as part of the guiding coalition.
Here are 4 steps you need to take to successful create the group that can guide this through:
Identify key stakeholders: you need a broad group, not only tech-savvy teachers. You should include teachers, leaders, students, parents and others from the local community that can support you.
Work with this group to develop a shared vision: this should align with the school’s mission and goals, and focus on the benefits AI can bring to students.
Use a variety of channels to communicate the shared vision to all stakeholders.
Empower the coalition to drive through the change process by giving them the resources, support and autonomy they need for success. It’s important to identify any training and resourcing needs early on so that you don’t come up against blocks. Do leverage the combined expertise and influence of the group: remember that you can’t do this alone.
3. Develop a vision and strategy
Kotter emphasises the importance of ensuring the school’s vision is clear and well-communicated. It is challenging to follow something that is hazy, and nothing erodes faith in leadership more than minds that keep changing every time an obstacle is hit or a shiny new innovation appears. As Kotter says:
In a change process, a good vision serves three important purposes. First, by clarifying the general direction for change, by saying the corporate equivalent of “we need to be south of here in a few years instead of where we are today,” it simplifies hundreds or thousands of more detailed decisions. Second, it motivates people to take action in the right direction, even if the initial steps are personally painful. Third, it helps coordinate the actions of different people, even thousands and thousands of individuals, in a remarkably fast and efficient way.
The importance of a vision in motivation cannot be underestimated. We need that clear goal to be heading towards to get us through the challenges that will invariably present themselves to us along the way. So, what does an effective vision look like? Kotter gives us 5 helpful indicators.
First of all, a good vision is imaginable. It should convey a concrete picture of ‘what the future will look like’. However, it has to be realistic, and grounded in the current realities of your school. There is little point in painting some future vision which encompasses moving the entire school to a project based learning curriculum and open plan learning spaces by next term if you’re stuck in a 1920s red brick school house and you’ve exam boards and Ofsted breathing down your neck. All that will achieve is you stressing out your team and making them question your grip on reality. And, as a leader, this is the fast track to failure.
Better instead to take the time to really understand where you are before you decide where you’re heading. There is nothing wrong with having a 5-year vision to achieve the above, but make sure you break this down into smaller, more achievable parts. You might therefore have a 1-year, 3-year and 5-year vision. Aim high, think big, but don’t be unrealistic about how quickly you can get there. We all know how slow-moving school systems can be.
Secondly, the vision should be desirable. It should point to a future state that is considerably better than the present reality. There is little point in a vision which does not improve the quality of both teaching and learning in a school.
The best way to consider this is to imagine the impact the change vision will have on stakeholders. I quite like these three questions, which I have adapted from Kotter and which are worth asking from the start:
If the vision is made real, how will it affect students? Will it improve their learning? Will it make them happier and healthier? Will it promote good behaviour and wise choices? In 5 years, will we be doing a better job than we are now and will student learning benefit as a result?
How will the vision affect parents? Will it keep them happy? Will it improve home/school relationships? Are we likely to provide better examination results and university destinations than we are currently? Will it make our school more attractive in the local market?
How will the vision affect employees? ‘If they are satisfied today, will this keep them happy? If they are disgruntled, will this help capture their hearts and minds? If we are successful, will we be able to offer better employment opportunities’ than other schools with whom we compete for the best teachers?
Thirdly, the vision should be feasible. The first thing the cynical physics teacher (and sorry for stereotyping here) will say is, ‘ok, but how…?’ Don’t fall into the trap of overpromising and underdelivering: it’s the worst mistake a new leader can make and can destroy your credibility. And, as Kotter says, “A vision that requires only a 3 percent improvement per year will never force the fundamental rethinking and change that are so often needed in rapidly shifting environments. But if transformation goals seem impossible, they will lack credibility and thus fail to motivate action.” It’s all about hitting the sweet spot: not so little change that people wonder what the point is, nor so much radical change that people think you’re crazy and aren’t prepared to go on that journey with you.
Finally, the vision should be focused, flexible, and easy to communicate. It must be “clear enough to provide guidance on decision-making”, “open ended enough to allow for individual initiative and for changing conditions”, and not “so complex that communicating it to large numbers of people is impossible”.
4. Communicate the change vision
In all the schools I work with, the single biggest determiner of success or failure is leaders’ ability to communicate the right message to the right people at the right time. It is remarkable how often we forget to do this: we meet as an SLT and come up with what we believe are well though-through ideas, then wonder why they don’t achieve the change we desire. It is nearly always because we haven’t been clear or consistent enough in how we communicate our vision.
Kotter outlines 7 elements to the effective communication of a vision for change. Follow these and you can’t go far wrong:
Simplicity: get rid of jargon, eduspeak, and all acronyms. Educators often hide behind language that we believe everyone will understand. Don’t make that mistake. Keep it simple.
Metaphor, analogy, and example: consider your audience and the mental pictures that will most resonate with them. Teachers are usually good at this, as we are used to teaching a class of 11-year-olds followed by a class of 18-year-olds.
Multiple forums: “Big meetings and small, memos and newspapers, formal and informal interaction—all are effective for spreading the word.” Consider also electronic media, and not only email.
Repetition: People need to hear the same message a number of times over before it will sink in. Don’t think you’ll bore or annoy your stakeholders: with a vision as big as yours, they’ll not grow tired of hearing it proving it is communicated with energy and passion.
Leadership by example: Practice what you preach. If you are extolling the virtues of ChatGPT, make sure you are using it and visibly showing its impact. This is easy for me, as it’s baked into my everyday practice. For example, writing this book has been so much quicker and easier because I can ask ChatGPT and Bing Chat questions rather than continually researching via Google.
Explanation of seeming inconsistencies: If you fail to address anything inconsistent you’ll undermine your message. If you insist on all your teachers using ChatGPT in their lessons but you continue to expect handwritten reports every term, you’ll have them asking where the time savings actually are. You don’t need to move wholesale over to AI, but at least consider what the most labour-intensive aspects of a teacher’s job are and think of how AI can help them.
Give-and-take: Listen, listen, listen. It’s not just about what you say: it’s about what you hear. Remember, you have one mouth but two ears. John Hattie gets it right when he talks about the best feedback not being what the teacher gives the student in the form of marks in their book, but rather the feedback the class gives the teacher on a moment by moment basis during a lesson. School leaders should therefore be at an advantage over other industries as we are primed to absorb feedback and alter our approach as a result.
5. Empower your team
As we saw in a previous post, one of the biggest barriers to accepting change is learned helplessness. Teachers often feel stuck in a system they can do nothing to change, so they give up, learn to fit in, and grumble pointlessly in the staffroom to anyone who will listen. As Henry David Thoreau once said, "The mass of men lead lives of quiet desperation and go to the grave with the song still in them." Let’s not allow that to happen in our schools as we bring in AI: let’s instead see this as a way to empower teachers and students to have significantly more agency and to create their own teaching and learning pathways with our support and guidance.
However, there are a number of what Kotter refers to as ‘structural barriers’ that need to be removed if staff are to have the agency and confidence to support your change agenda. He names four: “structures, skills, systems, and supervisors.”
Firstly, schools can be highly structured workplaces, with tight lines of accountability, a hundred different policies to adhere to, Ofsted standards to uphold, and examinations to work towards. How on earth to innovate within all that? Leaders need to consider how to loosen up a bit, perhaps not to expect the same outcome at the same time from the same students. Consider giving over a few days or a week to AI-supported project work, which will be low risk but high impact. If we believe enough that AI will significantly impact on the teaching and learning paradigm then our structures will have to change, moving from siloed subjects towards interdisciplinary, immersive, real-world experiences. However, that will not happen overnight, so consider which structures can be changed now, and which can be changed over time.
Secondly, teachers may not feel they have the skills to effectively drive forward these changes. They may feel lost, scared, and unsupported. This can significantly undermine action, leading to mutual frustration and resentment. ‘It’s all very well you wanting me to use Midjourney in my lessons, but I don’t even know what Discord is.’ New jargon, new techniques, and a hundred new AI solutions coming out every week: no wonder your 55-year-old history teacher looks lost. It goes back to an earlier point: you must keep things simple, and make sure you provide enough training so teachers feel comfortable. When you first introduce these tools don’t expect anyone to use them in the classroom for several weeks, maybe even more than a month. Allow everyone to get the feel of them in their own lives before considering applying them in lessons. Keep it lighthearted, low-key, and without expectation.
Thirdly, if you don’t have the right systems in place, your team will get frustrated very quickly and not bother to continue. The classic one here is internet speed. If your whole school is online at the same time, can your network and servers cope? The other blocker will be hardware: teachers who can’t access the most basic functionality because they have a 6-year-old Intel i3 laptop with a broken screen will hardly jump up and down with glee when you ask them to use ChatGPT in their teaching. Make sure you have the basics in place.
Finally, teams are always hyper aware of the Principal and what he or she thinks. As mentioned before, this is why, if you’re a keen teacher or Deputy who wants to lead these changes, you must ensure you get the Principal and Board of Governors fully on side. If the Principal comes into a lesson and sniffily says ‘I see you’re another lazy teacher who can’t be bothered to plan properly any more’ you’re in trouble. Either get them aligned, or get out and find a school where you can make a difference. I’m serious. You’ll never make long-lasting change if teachers are terrified of getting on the wrong side of the Principal by being too radical.
Equally, if you’re a new Principal in a school and you come across the recalcitrant Deputy who looks askance at you in every meeting and quietly undermines you with a few raised eyebrows and not-so-subtle remarks to colleagues (the sort that put ‘the future’ into air quotes), you must challenge them robustly and in private. Do not get into a heated debate with them in front of their colleagues, as you will invariably lose. People like this are looking for an audience: don’t stoop to their level. Instead, take the time to patiently listen to them telling you how they are right and you are wrong, before politely reminding them that it is your overall decision and that, whilst you respect their point of view, it’s not one you’ll be considering too deeply.
You cannot allow a few naysayers to undermine you: if you have taken the time to listen to stakeholders and collectively created a vision, then allowing a lone voice to undermine you will never get you where you want to go. If you have given them ample time to join your inner circle and contribute, and they have refused, then this is their problem not yours. And if they continue to cause you problems, perhaps this is not the right school for them any more.
6. Generate Short Term Wins
Quick wins are essential for gaining credibility. In The First 90 Days, Michael Watkins also stresses the importance of these, to create a positive impact, generate enthusiasm, and demonstrate your ability to deliver results. This isn’t difficult with AI: simply by showing teachers how ChatGPT can immediately save them time by supporting lesson planning you have immediately identified a major barrier to teacher efficiency and shown how AI can overcome it.
The easiest way to determine what these short term wins should be is to speak to people. The first stage in any change management process is to ensure you understand the challenges of others. It may sound harsh, but in general people only care about themselves: about their health, their relationships, and their finances. That is why all of marketing targets one of these drivers. Any resistance to whatever it is you may wish to introduce will have its root in one or more of these: perhaps your teachers are worried about working even harder which will impact on their health, or mean they won’t see their families as much, or perhaps they are worried you’re going to eventually replace them with an AI.
By meeting with your team and drilling down into the reasons for their resistance, you will achieve two results - first of all, to understand which levers to pull to get them on side, and secondly, to establish what these quick wins should be.
Let me give you one example. Say through your listening exercise you realise that one of the biggest concerns staff have over introducing AI is that it will make people lazy. One quick win will be to restructure the next staff training day so that you spend less time on any planning activities (as ChatGPT can blast through them faster than any human) and more time on fun, team-building activities.
If, by the end of the day, staff feel like they’ve got through the same amount of work but are considerably more motivated and refreshed, you can point to AI as being the reason for that (plus your brilliant planning). You can explain that, far from making them lazy, it simply allowed them to focus on more of what is important - and that that can make them more motivated than lazy.
Kotter points to three reasons why short term wins are important. First of all, they are visible: everyone can see that these are real results you’re getting from introducing AI. The staff have actually felt refreshed for the first time at the end of a training day. Who knew they could be such fun? Secondly, they are unambiguous: there can be little argument that AI had a tangible impact on their planning. Thirdly, they are “clearly related to the change effort”: if it can support teacher planning, freeing them up to do more of what they enjoy, then by extension the same can be applied to students in the classroom.
At this point it is worth referring back to Lewin, as he goes into a little more depth about the practical implementation of what he refers to as the change agent: in this case, AI. Once the existing state has been unfrozen, through your awareness raising and a collected understanding that there is no choice but to change, Stage 2 of his model focuses on introducing the change agent.
If teachers recognise how much of their time is taken in administration tasks, ChatGPT writing schemes of work, lesson plans and resources is a godsend. If they’ve struggled for years with creating differentiated learning materials, being able to subtly tweak the language or detail of a resource, or to turn written text into audio or video, will open new doors to serving a wider ability range, and will as a result make them better teachers.
If students can now input their work into an AI mentor and receive detailed, helpful feedback, they will soon see how they can progress at their own pace, not at the pace of the teacher who may take a week or more to assess work and hand it back. All of this is good, and all of it is available now.
However, unless we initially raise awareness of how broken the current system is, there will be less urgency to make the effort to change. As we all know, change can be hard work and it can be painful. This is why it is important to both clearly show how ineffective the system currently is, as well as having a vision for how to move us away from this and into a more positive future.
7. Consolidate Gains and Produce More Gains
Once you have the momentum in your favour through the impact of short term wins, you can use this to take on further change projects. It’s all about building one change onto another: from reducing staff workload to focusing their attention onto using the time saved to create new inquiry-led projects, and from empowering students to use AI in simple ways (such as supporting assessment) towards them taking more ownership to create their own learning resources based on where they are in their learning. Because of the rapid improvements in AI I believe that these gains will naturally keep consolidating as we are able to do more and more with it, and it becomes further automated and ubiquitous. It almost feels like, if we can introduce it and get everyone feeling comfortable with it, the rest will follow.
You may at some point need more help, perhaps employing experts in inquiry-led learning, or training key staff up to take the lead. As we saw earlier in the chapter, new roles are envisaged that move us away from leadership being all about accountability and towards expert support and the oversight of a new form of learning architecture. Design thinking will become increasingly important if AI can take away much of the need to hold others accountable.
The role of school leaders at this point in the change process is critical: it’s all about keeping the eye firmly on the vision and maintaining effort and urgency levels. It is all too easy for exam and inspection pressures to reassert themselves and for teachers to soon fall back into the old way of doing things. Without purposeful and enthusiastic leadership, drift is all too easy. This is why having a clear and concrete vision is so important: by continually reminding staff of where you’re headed, they can refocus on what you all know is most important.
Ensuring that there is some technical management is key. It’s not only about oversight, vision and constant encouragement: you also need boots on the ground organising and project managing. This is where your middle leaders come in. Assign your strongest middle leaders to lead on certain projects that drive forward your vision, such as collapsed curriculum ‘AI weeks’ and new assessment processes. By dropping the ownership of strands of your vision down to your middle leaders you are also sending a strong message, that you’re all in it together and that the success of the change process is a team effort.
Finally, think about what can be abandoned through the process. If AI enables you to analyse data in minutes, do you really need your middle leaders to provide you with a monthly report on progress? And if students can assess their work in real time with AI support, do you need to insist on books being fully marked once per week? There may be legitimate reasons why these are both still required, but make sure you don’t just do things ‘because they’ve always been done like that’. Look at how AI can remove some of the more tedious admin and free everyone up to be more balanced, energised and creative.
8. Anchor new approaches in the school’s culture
There is a well-worn management saying that is generally attributed to Peter Drucker: “Culture eats strategy for breakfast”. Kotter stresses the importance of culture for three main reasons:
Because individuals are selected and indoctrinated so well.
Because the culture exerts itself through the actions of hundreds or thousands of people.
Because all of this happens without much conscious intent and thus is difficult to challenge or even discuss.
The challenge of any leader bringing innovation into a school is to counter any prevailing culture and not graft “new practices” onto “old cultures”. If your staffroom is filled with teachers who have been in the same role for twenty or more years, and your parent body expect desks in rows and weekly spelling tests (and will take their children to another school if you get too radical), then going straight for the inquiry led, open plan approach is likely to end in tears (and you being removed by the Governors).
If, however, you start small, perhaps identifying one more forward thinking teacher to try out a few ideas with one class, word will soon start to spread. If the students are enthused, and tell their parents how much they learnt and how enjoyable it was, then this slow trickle may turn into a flood. Students will always be the most effective salespeople in your school.
Take your time. We’ve had this system for more than two hundred years and society hasn’t collapsed. It’s tried and tested. It may be horribly outdated in so many areas, but the worst thing you can do when you begin your change process is trash everything that has gone before. It doesn’t have to be zero-sum: you can pay tribute to tradition whilst gradually promoting a new way of working and learning. Don’t destroy the old: step by step show how the new is better-suited to the needs, attitudes and aptitudes of students in the 21st Century.
Here are a few steps you can take to help you start changing a school’s culture:
Talk about how evidence shows that this new way of AI-supported teaching and learning leads to greater engagement and better results.
Talk about where the old education culture came from, how well it has served us in the past (and why), but why we now need to change.
Have one to one conversations with any teachers who refuse to adapt to the new culture, showing them how AI make their teaching more interesting and can save them time for things that are most important to them. If they realise that you won’t be going back to the old way, they are likely to either comply, or leave/retire if the culture no longer fits with their beliefs (the cognitive dissonance we explored earlier).
Make sure that any new hires are thoroughly screened for fit into the new culture. This should come through job descriptions and interview questions.
Do not promote anyone who does not wholeheartedly promote the new school culture.
One thing to bear in mind is that culture change comes at the end of the process, not at the beginning. Don’t try to change a culture before you bring in your AI innovation: culture will only shift when it needs to in order to accommodate a new pedagogy. Once people can see that their teaching practice has become more efficient, lessons are more fun, students are more engaged and outcomes accelerate, culture change will follow without anything explicit needing to be done.
However, particularly in the early days, you must be talking the talk and walking the walk day in, day out. You cannot lapse. Promote, encourage, train, support, challenge - you need to push the agenda constantly and consistently. The radical shift that AI looks likely to create in our entire educational system is both exciting and terrifying, so we have to treat our colleagues with kindness and sensitivity as some are scared. Making anything compulsory, disciplining staff who don’t use ChatGPT in their lessons, is hugely counterproductive and likely to end in failure.
Having said that, sometimes you do need to let people go if they refuse or seek to undermine your efforts: and that can include the most senior people in your school. If you believe that bringing AI into your school is ultimately the right decision for the future of the students you educate, then you don’t want anyone stopping you from achieving this. Besides, it’s pretty miserable being stuck in a system you despise, so it’s usually for the benefit of both sides that you part company with those who won’t change to fit the new culture.
The final part of Lewin’s model is similar to Kotter’s stage 8. Once the changes have been introduced they need to be normalised. Moving from siloed lessons with students facing forwards to interdisciplinary, immersive real-world projects is a huge leap, and will take time to seem commonplace. Some teachers and classes will adapt quickly, whilst others will take longer. Subjects which lend themselves to more of a project-based approach, such as the sciences and humanities, may find themselves adapting quicker than more linear subjects such as maths and English. However, with the right approach and mindset, every subject can make the necessary move in the next few years.
Again, it is all about culture at this stage: once the change has been practically implemented, it will be the language used, the attitude towards old and new, and how the school presents itself both internally and externally that will determine the success or failure of the initiative. Successes should be openly and widely celebrated and student progress should be promoted at every opportunity. Ensure you continually listen: feedback should not only happen in the early listening stages. Be open to what people are telling you and don’t fall into the trap of being too emboldened by any early successes and therefore blind to the pitfalls that will inevitably appear along the way.
Building communities of support
Leading such seismic change is an enormous undertaking, and we cannot do it alone. It is vital that educational leaders the world over join to support one another. Running schools can be a lonely and isolating job, but it no longer needs to be. The online community is growing and getting stronger by the day. A new movement will form, one that privileges sharing and support over siloing and competition, and communities of practice rather than league tables and inspection results that define a school’s success based on a narrow set of criteria.
So my stage nine of the above process is to share good practice in worldwide communities of change leadership. Come follow me on LinkedIn, Twitter or Substack and connect with the thousands of educators and leaders who are already taking advantage of what social media has to offer. And in the same way that Martin Luther galvanised an entire religious movement on the back of the creation of the cheap, mass-printed book, so we as educators can create a movement that takes us from where we are now, to where we have to be. We owe it to those we teach.
This excerpt is taken from the book “Re.Generate: How Generative AI will Revolutionise Education”, available later this year.
Kotter, John, Leading Change, HBR Press, 2012
Lewin, Kurt, Frontiers in Group Dynamics, 1947
Watkins, Michael, The First 90 Days, HBR Press, 2013