How we’re exploring the role of AI
Early years professionals have one of the most crucial jobs imaginable: nurturing young children and setting the foundation for lifelong learning.
At Nesta, one of the core pillars of our fairer start mission1 is supporting the early years workforce. With artificial intelligence (AI) already on the rise in primary and secondary schools, we’re running a programme of work2 to explore how AI can empower early years professionals to provide more high-quality and effective early childhood education and care experiences.
This report outlines the three most promising matches we’ve found between what AI can do and what early years professionals need. We also share what we’ve learned about why there’s not more innovation already happening in the sector.
Building from these insights, we’re now starting to test three ideas that could bring the benefits of AI to early years professionals and address the barriers that prevent innovation from starting and spreading in the sector.
If you’d like to hear more, please get in touch3.
How we’re exploring the role of AI in Early Childhood Education and Care
We are working towards a vision of an Early Childhood Education and Care (ECEC) system that is enabled and empowered to deliver better childcare quality and outcomes for children, through AI. This programme will stress test this vision through experiments, explore roles Nesta could play to reach it, and/or build evidence on whether or not this vision is desirable, viable and feasible.
Our project consists of three phases, and this report shares key learnings from the first phase
- Understand: Identify and start to demonstrate the role Nesta could play to catalyse the use of AI to improve the efficiency and/or effectiveness of frontline practitioners
- Design: Deliver/enable interventions that people use/benefit from in the real world and create the desired results
- Scale: Grow interventions/intervention enablers so that AI tools can reach all the professionals across the UK who can benefit from them.
Through summer/autumn 2024, we ran a series of sprints to develop insights on the pain points in early years that could be amenable to AI tech solutions and understand the factors that affect the implementation of tools like this in nursery settings. Throughout the project so far, we have carried out numerous interviews, speaking to early years innovators, ECEC sector leaders, ECEC front-line professionals, academics and investors. We also undertook desk research including a review of relevant reports and research on ECEC and AI, as well as sector news.
What’s the current state of play for AI and technology in early years?
What’s the current state of play for AI and technology in early years?
Nesta’s recent Innovation Sweet Spots4 discovery work shone a light on the ever-evolving landscape of early years tech innovation, identifying a number of different areas of research & development (R&D), investment, growth and innovative solutions targeting the early years sector as a whole. Much of this work identified a number of opportunities that have a broader reach than just the early years (spilling out into broader EdTech territory), with the majority having a primary focus on children/young people or parent/caregiver-facing innovation. However, there were few examples that aimed at supporting professionals in their roles, or the wider childcare system, within the context of improving the quality of provision.
The ‘nursery tech’ sector, ie, innovation that supports professionals, settings and the system, represents a relatively small portion of the overall EdTech market, with our research suggesting around 300 start-ups spanning seven key categories: nursery management, parent engagement, finances, development assessment, workforce management, marketplaces and content and learning. AI adoption in this sector remains limited, with only a few companies integrating small AI features, such as assistants, chatbots, analytics and speech and recognition.
The work suggests that, although there is some growth and development in nursery tech, it is still small and a fraction of what its potential could be. Additionally, AI could have the potential to transform categories of nursery tech and problems to solve for practitioners, settings and the system, but this growth has not yet materialised.
Where can AI be applied to increase the quality of Early Childhood Education and Care?
Where can AI be applied to increase the quality of Early Childhood Education and Care?
We’ve identified three key areas in which the technological capabilities of AI could be applied to support early years professionals.
1. Enhance professional development for the ECEC workforce
The problem
Opportunities for CPD are limited and existing courses are of variable quality. Nurseries cannot afford expensive courses and it is hard to arrange cover for staff in order to attend courses. There is a large range in educational experiences amongst early years professionals - one size will not fit all.
Why AI?
There’s an opportunity to improve CPD through 1) personalisation, 2) embedding real-time feedback/coaching and 3) making CPD available in short downtime moments.
Outcomes we want to see
Professionals have better access to skills and knowledge resulting in higher quality interactions with children. Nurseries achieve a greater impact on practice for the same budget compared to existing training.
2. Reduce administrative burden in order to increase the time professionals spend on activities that directly benefit children’s development
The problem
Nursery managers experience a high burden of administrative tasks, reducing the time available to support and coach their staff and enhance developmental practice. Tasks include record keeping on children and staff, regulatory reporting, financial management and funding applications.
Why AI?
Automation of admin tasks could reduce workload. Many tasks require synthesis of information and report writing which is highly amenable to generative AI.
Outcomes we want to see
A reduction in manual admin tasks for practitioners, resulting in increased time to work directly with staff and children.
3. Enhance the quality of assessment and developmental activities for children, through AI-powered personalised recommendations for practitioners
The problem
Understanding a child’s developmental stage and the activities and support they require to advance is critical to quality provision, and this requires a very high level of knowledge, skill and creativity. Personalisation and tailoring to the individual needs of children is time-consuming and challenging.
Why AI?
AI has the potential to bring information about a child together to give an accurate picture of their development stage across multiple domains, and combine information about their preferences and best practice to give creative ideas for developmentally appropriate activities.
Outcomes we want to see
Development of personalised child development plans, tailored to individual needs resulting in meaningful and developmentally supportive interactions between nursery professionals and children.
What are the barriers to innovation, development and adoption of new and effective technologies?
What are the barriers to innovation, development and adoption of new and effective technologies?
1. The early years system is stretched with little room to stop and experiment
The ECEC system faces significant financial constraints, high workloads for practitioners, and challenges in attracting and retaining qualified staff. These factors leave little room to stop and improve, with nurseries prioritising immediate needs over exploring innovative ways to improve outcomes (eg, new technologies).
The ECEC system in the UK faces significant challenges, including financial constraints, high workloads for practitioners, and difficulties in attracting and retaining qualified staff. These pressures result in nurseries prioritising meeting their core responsibilities over innovation or testing new products and services. Nurseries operate on tight margins, making it difficult to invest in new technologies.
Government funding often does not cover the full cost of nursery places, and funding models do not account for administrative time, so practitioners have very little time outside of their child-facing roles in which to explore or implement new approaches. High staff turnover, with rates as high as 40% in some settings, creates a constant need for training, further limiting the capacity for experimentation and innovation.
2. Tech readiness across ECEC is low
Many ECEC settings exhibit low levels of technological readiness, stemming from a combination of factors such as poor physical infrastructure and digital skills. New tools need to integrate as seamlessly as possible into current ways of working.
We heard that ECEC practitioners may be sceptical of new technologies for several reasons including fatigue from previous initiatives that failed to deliver a positive impact, and awareness that there are limited resources for successful tech integration.
Early years settings do not consistently have access to reliable hardware and connectivity. We heard that limited broadband access and outdated hardware such as laptops and tablets are common issues.
3. ECEC is a tough market for innovators to operate, especially if the ambition is to reach disadvantaged families
The UK nursery market is highly fragmented, with the majority of nurseries being small- or medium-sized enterprises. This fragmentation makes it difficult for innovators to achieve scale with new technologies. Although large nursery chains exist these often cater for families with higher income.
Innovators face challenges in accessing ECEC settings for testing and gathering feedback on AI solutions, and investment in new technology-based solutions for early years education is limited. Understandably, given the standards they adhere to and limited time for experimentation, nurseries may demand robust evidence of benefits before considering new technologies, which limits opportunities for testing and improving new solutions.
Tech innovation targeting early years settings is considered less attractive for venture investment, primarily due to the fragmented market and low purchasing power of nurseries.
4. Deep tech innovation requires expensive and scarce expertise
There is a concern that innovators in the early years space could lack the expertise to effectively leverage AI, limiting the impact of AI-driven solutions. Whilst innovators don’t necessarily need deep AI expertise, they do require the ability to assess different models for their specific use cases.
This barrier may not necessarily be universal across the early years development space. Innovators have successfully deployed AI without needing advanced skills by relying on existing AI models and tools, often described as ‘wrappers’. However, such tools are relatively easily replicated and therefore less attractive for investment. For more complex AI applications, particularly those requiring custom models, the lack of AI expertise can be a significant barrier.
The early years workforce holds a deep understanding of the problems that need addressing, but there aren’t opportunities for practitioners who have the ideas and aptitude to create new solutions to get the support and funding they’d need to bring their ideas to reality.
5. There are limited well-established ways to determine whether new AI solutions are trustworthy and safe
The lack of clear regulation around AI tools in early years education can create levels of uncertainty for both practitioners and innovators. We heard that concerns around trustworthiness and regulation of child-facing products were much higher than for practitioner-facing products, which contributes to the rationale for our focus on practitioner-facing interventions.
The innovators we spoke to did not see regulation (or lack of regulation) as a barrier. We heard that ECEC settings tend to group AI with any other tech and subject it to the same standards, this means they look for trustworthiness, safeguarding and GDPR compliance, and were satisfied with this approach.
We observed that when the issue of AI in the early years is raised, practitioners' concerns associated with AI tech interacting directly with children are brought to the fore and influence the initial reaction until they are reassured about how risks are mitigated or that children are not the intended audience.
What's next?
What's next?
We’ve developed a range of ideas and opportunities that respond to the insights and learning above. These ideas span the different roles Nesta could play5 to galvanise AI development for early years, from building new products to enabling practitioners and developers to innovate. We aim to take these ideas forward, using them as vehicles to learn about the best way to create direct impact or enabling environments that allow impactful innovation for early years to thrive. Our starter spaces include:
Empowering professionals through AI-enhanced, tailored learning
Imagine bite-sized, AI-powered training sessions designed to fit seamlessly into a nursery professional's day. This opportunity explores a CPD platform that offers personalised, scenario-based learning, empowering professionals to tackle real work challenges and improve outcomes for children.
Brokering relationships between innovators and early years settings
A matchmaking service that connects early years settings with cutting-edge tech innovators. By enabling collaborative testing and piloting of AI solutions, this brokerage fosters partnerships that address real challenges across the sector and helps build an enabling environment for impactful innovation to thrive.
Delivering a programme of innovation support for enterprising early years professionals
A development programme that supports early years professionals who have deep expertise in the problems to solve and an entrepreneurial ambition to create and scale new solutions. By providing product development and commercial expertise, this programme aims to increase tech/AI skills in the workforce and turn practitioners’ ideas into impactful, scalable solutions.
Work with us
In order to create the lasting impact we’re striving for, we can’t do this alone, so we’re looking for people from a variety of different backgrounds, viewpoints and perspectives to join us in building and testing these innovations. Additionally - the above starter spaces represent our initial bets for how we could create impact, but we know that they could be wrong and that there is a multitude of other impactful ways we could create lasting change in this space - so we need others to come along the journey with us.
- Co-develop – are you a developer, entrepreneur or early- years practitioner who’s looking for partners to build new or scale existing AI or nursery tech solutions, or do you have an idea but not sure what to do with it? We'd love to partner with you.
- Collaborate – are you a researcher, innovator or early years professional who has a vested interest in creating impact and learning that supports the use of technology to improve outcomes for children across early childhood?
- Convene – are you all or any of the above? We’re creating a network of like-minded people who care about innovating in early years and are looking to move this agenda forward.
If any of the above sounds of interest, and you want to be part of this journey, please get in touch6.