GenAI is now in our childcare centres. But there isn’t any guidance

GenAI has entered early childhood education and care (ECEC) faster than the sector has been able to respond. Educators are already using generic tools to draft reflections, write newsletters, organise planning ideas, develop policy language and make sense of documentation. Some services are creating their own rules. Others are still working out where to begin. The tools have arrived before the guidance.

That is the problem we need to take seriously. The question is no longer whether early childhood educators will use GenAI. Many already are, or soon will be. The more urgent question is whether they are being given the right support to use it safely, ethically and in ways that align with Australian early childhood practice. These are the questions at the centre of my recent article on GenAI-supported reflective practice in early childhood education and care.

Added pressure

At the moment, too much of this responsibility is being pushed onto individual educators, directors and providers. They are being asked to make decisions about privacy, consent, bias, accuracy, documentation and professional judgement while also managing staffing, compliance, inclusion, family communication, planning and quality improvement. These demands do not pause because a new technology has arrived. Asking each service to solve GenAI on its own is not sector readiness, it is added pressure.

This is why we need to be careful about treating GenAI as just another helpful tool. In some ways, it may be helpful. It can support educators to move past a blank page, draft clearer communication or organise messy thoughts at the end of a long day. In a sector where workload and burnout are real, those possibilities matter. But support without guidance is not enough. 

Generic tools are not neutral

The issue is that many educators are using generic GenAI tools that were not designed for Australian ECEC. These tools may not understand the Early Years Learning Framework in any meaningful way. They may not reflect the National Quality Framework, the importance of play-based learning, family partnerships, cultural responsiveness, children’s rights, or the relational nature of early childhood pedagogy. They may produce language that sounds professional while missing the values that sit underneath the work.

That matters because the EYLF is not just a document educators mention in planning. It represents years of thinking, consultation and sector knowledge about what early childhood education should value in Australia. It foregrounds belonging, being and becoming, and asks educators to think about relationships, identity, culture, place, play, equity and respect. If GenAI tools are not grounded in that context, their outputs may sound polished but still be misaligned.

No alarm bell

This is where the risk becomes more than technical. It becomes ethical. The early childhood sector has worked hard to centre voices and perspectives that have too often been underrepresented. This includes Aboriginal and Torres Strait Islander perspectives, children’s voices, family knowledge, inclusive practice and culturally responsive ways of working. If generic GenAI tools reproduce dominant assumptions, flatten complexity or ignore context, those voices may not be strengthened. They may be quietly buried.

That does not happen with a loud alarm bell. It happens through small, ordinary decisions. A learning story sounds fluent but loses the child’s voice. A reflection uses the right words but misses the family context. A planning idea looks efficient but narrows the possibilities for play. A policy draft sounds sensible but does not reflect the rights and realities of very young children. The danger is not always that GenAI produces something obviously wrong. Sometimes the danger is that it produces something that sounds right enough.

Children’s data is not just another input

Privacy adds another layer. Young children cannot meaningfully consent to their information being entered into a GenAI tool. They cannot ask where their data goes. They cannot question whether an AI-generated reflection represents them fairly. Families may not even know that these tools are being used. If educators are entering children’s words, images, behaviour notes, learning records or family circumstances into generic tools, then we need a serious conversation about children’s digital rights.

Families must be part of that conversation. Some may be comfortable with GenAI being used to draft a general newsletter or staff meeting agenda. They may feel very differently about tools being used to support individual learning stories, reflections or documentation linked to their child. Services need clear ways to talk with families about what is being used, why it is being used, what information is protected and where the boundaries are. Trust cannot be assumed. It has to be built.

This is not a simple problem. I read, write and think about GenAI constantly in my role as an academic and researcher, and I still find the ethical questions complex. The technology changes quickly. The research is still emerging. The policy environment is still catching up. If this is complex for people immersed in the field, we cannot expect every educator or director to solve it alone while already under immense pressure.

This cannot be solved service by service

So what needs to happen? We need sector-wide guidance that is specific to ECEC. And we need practical policies that help services make decisions without starting from scratch. We need professional learning that respects educators’ time and explains not only how to use GenAI, but when not to use it. We need clear privacy guidance, family consultation, approved systems where possible, and shared language around consent, documentation and professional responsibility.

GenAI in early childhood cannot be shaped by technology companies alone, or solved service by service by already stretched directors. It needs input from educators, families, researchers, policy makers, peak bodies, Aboriginal and Torres Strait Islander representatives, privacy experts and children’s rights advocates. This is not just a technology issue. It is about making sure early childhood values are built into the way GenAI is used.

The goal should not be to shut GenAI out of ECEC. That is no longer where we are. These tools are already being used, often without clear guidance, and they may bring real benefits for workload and communication. The task now is to move from individual experimentation to shared responsibility. We cannot leave that decision to overloaded educators and directors working it out service by service. Either the sector shapes GenAI use, or generic tools will quietly shape practice for us. The sector needs a coordinated response that supports educators, respects families and keeps children at the centre.

Author

Nicole Thompson is a lecturer in early childhood education at Southern Cross University. Her research focuses on the ethical use of GenAI in early childhood education and care, with particular attention to children’s rights, data privacy, professional judgement and sector-specific guidance.

This article was originally published on EduResearch Matters. Read the original article GenAI is now in our childcare centres. But there isn’t any guidance here.