Why the best AI adoption strategy is a people strategy

Most schools have an AI policy. Far fewer have changed how teachers work.

AI adoption globally is outpacing even PC and internet adoption, with generative AI reaching 53% of the population in just three years. Now, 88% of organisations have embedded AI in their workforce, and 4 in 5 university students regularly use GenAI.

Classrooms across Australia are no different. But keeping pace can feel incredibly challenging, particularly when teachers already have full workloads. Atomi’s recent Tech in Schools 2025 survey found 25% of educators say time is the biggest barrier to integrating new tech, rather than a resistance to AI itself.

Getting AI adoption in schools right is complex. You’re balancing the need to protect students and safeguard academic integrity, while also exploring how AI can free teachers to work more effectively and improve student outcomes.

In conversation with Caitlyn Mossie (Head of Customer Success and Growth at Atomi), this blog unpacks why AI adoption in schools is actually about change management, why a student-focused AI policy isn’t enough and practical tips for where to start with your school’s AI rollout.

See the full interview with Caitlyn on YouTube

Why is AI adoption about change management, not just technology?

On the surface, the hardest part of a school’s AI rollout can seem like the technology itself. But in Caitlyn’s experience, the bigger challenge is managing so many different stakeholders.

From parents and students to teachers, senior leadership and heads of departments, each group brings a unique perspective and set of interests that need to be considered.

The schools that manage AI adoption best are those with strong change management principles in place. That means clearly communicating the “why” behind the change, bringing all stakeholders on the journey and gathering feedback to ensure the process feels collaborative (rather than a top-down mandate).

“The AI space is just as much about change management as it is about technology. Schools that come back to good change management principles are the ones that are really going to set themselves up for success.”
— Caitlyn Mossie

How do the best schools hold the tension between opportunity and risk with AI?

AI in education is a hotly debated topic, and with good reason.

On one side of the debate, you have the question of how this fast-moving technology is impacting students’ ability to think critically and problem-solve. There are also concerns about the risks posed by AI to data privacy and academic integrity. Schools are often focused on developing policies that define acceptable vs. unacceptable AI use, particularly regarding plagiarism and cheating in assessments and exams.

On the other side, you have the opportunities AI presents for both teachers and students. Not only are there immediate wins in reducing workload and accelerating what teachers can do, but AI use also has the potential to improve student outcomes (enabling differentiation at scale and the delivery of personalised, real-time feedback).

Caitlyn shares that schools who are successfully adopting AI are holding both realities at the same time. Educators are being pragmatic about the potential risks of AI, while ensuring it doesn’t stop them from exploring how this new technology can help them deliver great teaching more effectively.

“They’re not ignoring the risks that AI has for students and they’re not pretending they’re not there. But they’re also not letting that hold them back from thinking about how AI can help their staff.”
— Caitlyn Mossie

Why isn’t an AI policy enough on its own?

In Caitlyn’s experience, the AI policies produced by many schools are really student policies (focused on compliance and appropriate AI usage). The missing piece is to build a policy that answers the question: “What could teachers be doing differently with the support of AI?”

“A lot of schools are starting with AI policies as they relate to students: compliance, risk, plagiarism, academic integrity. But you can miss half the picture when that’s all you’re focused on. What some schools miss is the opportunity for staff and teachers to make their teaching better and make it easier to reduce workload.”
— Caitlyn Mossie

The shift from compliance to capability shows up in small, specific ways:

Compliance-first policy approach

• AI policy is written for students, focused on academic integrity, plagiarism and assessment risk
• Conversation starts with “how do we manage AI?”
• Rules are set top-down and communicated once
• Trying something new and getting it wrong is a risk to be managed
• AI adoption is treated as a technology rollout

Capability-first policy approach

• AI policy sits alongside a plan for staff: what could teachers be doing differently?
• Conversation starts with “how do we make our staff more effective?”
• Curious teachers are made visible and peer credibility does the convincing 
• Trying something new and getting it wrong is expected and an opportunity for learning 
• AI adoption is treated as a change management process

There is always a place for setting clear expectations for how learners can use AI. But the shift that needs to happen is moving from only considering “how do we manage the risks of AI?” to “how do we also make our staff more effective with AI?”

The reality is that students will live and work in an AI-shaped world. The question isn’t whether they’ll encounter it; it’s whether they leave school knowing how to think alongside it, not just use it.

Practical tips for where to start with your school’s AI rollout

Getting started with your school’s AI adoption doesn’t have to be a huge undertaking. Instead, a few small, deliberate moves can make a massive difference in your rollout plan:

• Start with a conversation, not a tool: Ask staff questions like “What’s taking up time that shouldn’t?” or “Where’s energy going that never reaches a student?” This gives teachers ownership of the problem before any solution is presented to them.
• Protect a small pocket of time for AI experimentation: Teaching can be incredibly reactive; an unexpected problem can easily override a well-planned week. Consider setting aside 20 minutes in an existing staff meeting, or a single protected slot each term, to explore how AI can be used in the classroom. It’s enough to get momentum going without asking anyone to find hours they don’t have.
• Let the willing teachers go first and make it visible: You don’t need everyone on board at once. Find the staff already curious, give them space to try something small and let their results do the convincing. Peer credibility moves people faster than a mandate.

“One place we really recommend you start is having really productive conversations with your staff, asking questions that get a vision out there. Consider: ‘What would really good teaching look like if we could make your lives easier with AI?’ Coming back to those core principles and asking ‘what if’ with people who really care is a really good place to start.”
— Caitlyn Mossie

Getting AI right starts with people, not a policy

The schools Caitlyn sees getting AI right aren’t the ones with the most detailed policy document. They’re the ones treating AI adoption as a change management process that has to hold opportunity and risk at the same time, and that starts with a conversation rather than a rollout.

That’s the same principle behind how Atomi approaches AI in the classroom: built around what genuinely improves teaching and learning.

If you’re thinking about where to start those conversations with your own staff, explore how Atomi approaches AI in education.