
Australian schools have absorbed more change in the past few years than in most of the decade before it.
Generative AI went from novelty to staffroom reality almost overnight. Cyber threats are growing more sophisticated, and the bar for child safety and data privacy is now higher than it’s ever been.
Meanwhile, curriculum and policy changes keep landing on Principals’ desks, creating more work for a profession that is already pushed to the brink.
It could be argued that much of what served schools well even five years ago is now being tested.
On Monday 16 November 2026, the implications of that challenge were explored at a roundtable hosted by ASX-listed tech firm Atturra.
The ‘Australia’s Education Future – A Tech Perspective’ roundtable gathered executives from the Catholic Education Network (CEnet), Microsoft and IBRS to discuss how shifting student and parent expectations reshape K–12 IT, from ethical AI and cyber safety to data and staff retention.
These demands tend to fall on the same few people, often a business manager or IT coordinator whose knowledge is largely undocumented. When they burn out or leave, that know-how goes with them, and the pressure is passed on to teachers and Principals.
How one school network saved $2m
Dr Joseph Sweeney, an Intelligent Business Research Services (IBRS) adviser specialising in workforce transformation and end-user computing, told The Educator that standardising support across a school network is the most reliable way to stop operations depending on a handful of key staff.
“I recently did a case study on Catholic Education Sandhurst Limited which standardised the device fleet with rapid provisioning [Microsoft devices with Autopilot], having additional replacement units ready in each school, and then bringing all support into a centrally managed service,” he said.
“When an issue arises with a device, it is simply swapped out for a spare and if needed, a tech is sent out. In Sandhurst’s case, this resulted in mid-sized schools recouping between $75,000 and $150,000 annually in local and diverse support fees, with a network-wide saving conservatively estimated at $2m annually.”
Dr Sweeney said Principals confirmed service levels had improved, though the model relies on unhurried staff training and remotely managed networks.
“In addition to the significant financial savings, the benefits of the Sandhurst approach are that it institutionalises all knowledge, and allows a fix at one school to be automatically applied across the entire school network where appropriate,” he said. “The standardise and centralise support model does not negate the value of the digital heroes in schools.”
Instead, Dr Sweeney said, it frees them to focus on more meaningful work – most importantly, leading the appropriate use of technology in the classroom.
Keeping teachers in charge of AI
As schools debate guardrails and digital detoxing, much of the AI conversation has centred on how much students and teachers should use it. For Lee Barrett, AI Education & Advocacy Lead at CEnet, that debate risks overlooking a more fundamental question: who is making the call.
“An AI tool can be used sparingly and still quietly label a child, and it can be used every day without ever doing so; the difference comes from design and leadership, not from screen time,” Barrett told The Educator.
CEnet’s cechat platform for Australian Catholic schools, built on the University of Sydney’s Cogniti code base, reflects that thinking: its AI agents are configured by educators and act only when someone engages with them.
Barrett said an early version of CEnet’s inclusive education planning agent produced convincing descriptions of student needs that weren’t grounded in anything a teacher had observed.
“We rebuilt it so that it won’t describe a student’s needs without evidence from the teacher,” he said. “If the evidence isn’t there, it asks for it. The agent’s job is to sharpen a teacher’s understanding, not to fill gaps with a plausible guess.”
His advice to Principals is threefold.
“First, ask of any tool, ‘Who decides?’ If the answer isn’t a teacher, push back,” he said. “Second, train staff in judgement rather than in buttons. Knowing when to trust an output, when to check it and when to set it aside matters more than knowing where to click.”
Third, Barrett said, Principals should be wary of anything that profiles students or promises personalised insights about them.
“We want students doing the thinking inside learning experiences their teachers have designed,” he said. “AI should help a teacher notice more about a child. It should never do the noticing for them.”
Safe AI starts with school data
Brett Auton, Director of Scholarion, a K–12 school management platform developed by Atturra, said the tension between protecting data and gaining value from AI is real, but schools don’t necessarily have to choose one over the other.
“Schools that excel in this area see it not as a trade-off, but as a design decision,” Auton told The Educator. “AI should access the data rather than the data being sent out to the AI.”
Auton said that approach starts with the data itself, not the tools.
“One of the biggest risks isn’t a sophisticated cyber-attack; it’s a staff member downloading student information into a spreadsheet and leaving it in a Downloads folder or on a shared drive,” he said.
He added that many schools also rely on a patchwork of third-party software, some of it built by small vendors without the security or governance expected of a system that holds student records and wellbeing notes.
Auton said an AI tool’s safety is “directly tied to the integrity of the systems it draws from.”
“If it connects to a weak system, it inherits that vulnerability and may expose existing problems. So, ensuring solid data foundations matters more than deciding which AI tool to adopt next,” he said.
“When data is fragmented across disconnected systems, spreadsheets, and inboxes, each new AI tool adds another copy, integration, and risk factor.”
When data is unified, however, Auton said AI capabilities can be activated within a secure framework and inherit its existing protections.
“So, if a Principal is considering activating a new AI tool tomorrow, the crucial question they should ask first is: ‘What data could this tool access, and how will we control and protect it?’”
Where does your student data go?
Vivek Puthucode, General Manager at Microsoft Elevate Asia, said the key to balancing safety and innovation is getting the order right
“Start with the data: know what student information you hold, who can access it, and make sure the tool respects those permissions,” Puthucode told The Educator, adding that principals need to know where student data goes and who can see it.
“If the vendor can't answer that clearly – including whether the data is used to train their models – the tool isn't ready for your school. Then start small, with staff first, and build confidence before scaling to students.”