Gen-AI in early childhood education: Navigating a not-so-new frontier?

Have you had that experience where a colleague shares a topical link, and the group chat goes wild? Constructive and rich discussion filled with critical reflection follows and then slowly it fizzles out, lost to the group chat ether forever. Recently, we had that exact experience when a provocation about generative artificial intelligence (Gen-AI) in early childhood assessment was shared between us. We decided, however, that the conversation was too important to let slide. The landscape of early childhood education is rapidly evolving, with Gen-AI becoming well established as a powerful yet controversial tool. As educators, we find ourselves at a crossroads, grappling with the potential benefits and challenges of integrating Gen-AI into our assessment practices and daily routines. 

On one hand, Gen-AI promises enhanced efficiency in the documentation of the early years planning cycle. Educators experimenting with tools like ChatGPT report assistance with grammar, structure, and refining learning outcomes. This can be particularly beneficial for those who speak English as a second language or are new to observation-writing. 

However, concerns abound. There’s a fear that Gen-AI might over-interpret observational evidence, overstate child learning or make inaccurate assumptions. The risk of losing the nuanced, professional judgment that qualified educators bring to assessment is real. As one colleague pointed out, ‘Outsourcing the analysis of observation evidence to AI reduces professional decision making in a time when we are striving to raise the professionalism of the early childhood workforce.’ 

Ethical considerations further complicate the picture. Data privacy, the environmental impact of increased data storage and the potential erosion of trust between educators and families if AI-generated content is misused are all valid concerns. A poignant example shared in our discussion involved a parent questioning the authenticity of an observation of their child, noticing the language provided as evidence of learning did not match what the parent knew to be the child’s current ability. This highlights the delicate balance of trust we must maintain when using Gen-AI in assessment for and of learning. 

Despite these challenges, the reality is that Gen-AI has already entered our field. As one colleague noted, ‘The horse has bolted.’ Many early childhood educators are already using Gen-AI tools, often without formal guidance or consideration of the ethical implications. This underscores the urgent need for AI literacy in our profession. 

Developing AI literacy, however, is not without its hurdles. Many educators report low confidence in understanding and using AI¹, coupled with time constraints that make skill development challenging. The rapid pace of Gen-AI adoption means we’re often playing catch-up, struggling to develop guidelines and best practices. 

So, where do we go from here? 

First, we need to prioritise AI literacy in teacher education programs and ongoing professional development². This isn’t just about learning to use Gen-AI tools, but understanding their implications, limitations and ethical considerations. 

Second, clear guidelines and policies for the use of Gen-AI, and AI more broadly, in early childhood settings are essential. These should balance the potential benefits of Gen-AI with the need to maintain professional judgement and the human element of our work. As holders of professional knowledge and judgement, these guidelines should be developed collaboratively with both educators and academics in addition to policymakers, families and, where possible, the children themselves.  

Third, we must engage in ongoing discussion and research about Gen-AI’s role in our field. This includes exploring how Gen-AI might reshape the educator’s role and addressing the potential widening gap between those with and without AI access. 

Lastly, as we navigate the murky waters of AI use in our field, we must stay true to the core values of early childhood education. Gen-AI should be a tool that enhances our ability to support children’s learning and development, not a replacement for the critical thinking, creativity and empathy that define our profession. 

The integration of Gen-AI in early childhood education is inevitable, but how we shape its use is up to us. By approaching Gen-AI thoughtfully, ethically and with a commitment to ongoing learning, we can harness its potential while preserving the heart of what we do—nurturing the growth and development of young children. 

We aim to keep the conversation and critical reflection going: How are you using Gen-AI in your work with children?  

Cristina Guarrella, The University of Melbourne 

Caroline Cohrssen, University of New England  

Parian Madanipour, Griffith University 

The authors acknowledge their deliberate use of SparkAI, a closed, secure Gen-AI system in writing this blog to demonstrate the benefits of Gen-AI. A transcript of the group chat was uploaded to SparkAI, and two prompts were used to 1) identify key points from the discussion and 2) create an initial draft of the blog. Next, contextual information was added throughout the draft by the first author before further revision and editing by all authors to improve accuracy. Finally, references to research discussed in the group chat were added as footnotes.