Summary of International Trends in AI and Education

In recent years, the world has witnessed the rapid emergence of artificial intelligence (AI) as a transformative technology with profound economic and social implications. Nations across the globe are recognising the potential of AI and its diverse applications, ranging from business and healthcare to transportation and education. The education system is instrumental in the use and adoption of the technology, as well as to develop talent in the sector, so that the benefits of AI can be shared across society. The education sector has, and will need to continue, to adapt to the changing context of this century through considered implementation of emerging digital technologies, especially AI.

This summary outlines the key themes of the global AI strategy and policy landscape, with a particular focus on the education space.

        1. Leading AI Development

A notable trend in the landscape of national AI policies is the competitive race among nations to establish themselves as global leaders in AI. Countries such as the UK, Australia, Finland, and Singapore are actively seeking to leverage AI’s transformative potential for economic growth and are aiming to lead the implementation and development of AI. New Zealand has a thriving commercial AI landscape with many organisations adopting AI systems for data processing, automating business processes, improving customer relationship management and to support decision making. AI in educational technology (EdTech) will also be a growing area, New Zealand has many EdTech startups and companies, including Education Perfect which has been used in over 60 countries and has previously partnered with New Zealand Parliament to develop modules on civics education.

        2. Talent Development

The growth of AI technology demands not only on developers and engineers, but also experts from a wide range of backgrounds and disciplines who can ensure its safety and beneficial outcomes when it is applied. Around the world there is a scarcity of AI talent and an AI skills gap in workplaces, therefore upskilling the current and future workforce will be necessary. Developing and retaining AI talent is a critical component of effective implementation, and nations are focused on investing in education and training programs to meet this need. Australia, for example, has set up grant programs and scholarships for the development of AI solutions to regional problems and industry focused research projects. The EU is supporting digital competency in the education sector with the development of a digital competency framework for educators and free web-based self-reflection tools (SELFIE). New Zealand has a Digital Tech Industry Transformation Plan which has a focus on enhancing the skills and talent pipeline. The Digital Technologies and Hangarau Matihiko (DT&HM) curriculum was introduced in 2017 but AI literacy is not specifically included and educators may need more support in the sustainable implementation of technology into their teaching practice, as well as using and understanding AI.

        3. National Centres of Expertise

Countries have established, or are establishing, national centres of expertise that bring together researchers, businesses, and governments (e.g., UK’s Alan Turing Institute, Finnish Centre for AI). These centres serve as hubs for the adoption and development of AI technology. They foster collaboration, innovation, and knowledge-sharing, enabling nations to stay at the cutting edge of AI advancements and develop local products. Cross-disciplinary AI research is occurring across New Zealand universities, but increased investment towards a coordinated approach, and potentially establishing a national centre, could help to strengthen the research to application pipeline. In education, collaborative hubs have been set up to facilitate communities of practice between teachers and researchers, such as the European Digital Education Hub.

        4. Ethical Principles

The growth of AI technologies also brings a set of concerns, including the potential exacerbation of inequality, the increased spread of dis- and misinformation, issues of inexplicability and bias, and data privacy challenges. Addressing these concerns is essential for the responsible adoption of AI, and as AI technology advances, ethical considerations have come to the forefront. Principles are emerging related to transparency of use, testing for fairness and non-bias, human control and oversight, maintaining privacy, and accountability of tech companies. These ethical principles are crucial for ensuring the responsible and equitable deployment of AI in all domains and particularly education. The Institute for Ethical AI in education states that all learners should benefit from the use of AI in education; AI systems should promote equity between different groups of learners and increase the autonomy of learners, humans should retain oversight and responsibility for decision making and a balance will need to be struck between privacy and legitimate use of data to achieve desirable education goals. The European Commission has put out ethical guidelines for the use of AI and data in teaching and learning for educators. Australia has also drafted a framework for schools for guidance on the ethical and safe use of AI, which will also be needed in the New Zealand context.

        5. Assessment and Content

The education sector has recently been responding to the ready availability of publicly available generative AI (e.g. ChatGPT, DallE).  The integrity of assessment, certifications and qualifications will need to be upheld; maintaining the authenticity of individual work is a challenge. Strategies for improving assessment design to manage the access to AI include: asking students to demonstrate evaluative judgement, incorporating authentic, context-specific or personal assignments, demonstrating the learning process (e.g. checkpoints throughout a learning program), group-work & feedback or in-class discussions and also changing formats (e.g. video, podcast, website, pen & paper). A focus on promoting academic integrity and updating policy and procedures to respond to misconduct will be important at a secondary and tertiary level.

Projects around online hubs for digital content, particularly in higher education (e.g. Ontario, Canada’s eCampus, Finland’s Digivision 2030), are emerging to enhance the student learning experience. Massive open online courses (MOOCs) have been developed to educate the wider public about AI such as the University of Helsinki in Finland’s ‘Elements of AI‘ course and the University of Adelaide’s suite of MOOCs to support Australian teachers to implement the STEM and digital technologies curriculum with a course on ‘Teaching AI in the Classroom’ currently under development. Singapore is pioneering the use of AI in their national digital education platform, implementing personalised recommendation algorithms for student learning at the primary school math level. New Zealand Curriculum Online currently provides digital resources and assessments, though there is a wide scope for extending the platform and better integration for users in education settings. Digital presence and content will also be important for the New Zealand international education strategy.

        6. Diversity and Inclusivity

Adopting AI technology requires an interdisciplinary approach involving diverse perspectives to ensure benefits are widely distributed. Governance of AI is a recent area of global discussion which could include assessment of risks and benefits, developing best practices through research and collaboration, licensing around data or computing power and auditing the outcomes of algorithms and AI. The UK has set up an AI standards hub and governance centres which aim to ensure the responsible and standardised use of AI technologies.

New Zealand’s education strategy places learners at the centre, prioritising safe, inclusive spaces and reducing barriers. A focus on equity and learners’ needs can help ensure that AI integration in education settings will benefit all students, irrespective of their background. AI models have the potential to homogenize information and knowledge, occluding the perspectives and voices of minority groups. The data used to inform the AI models should ideally be representative of the population using or affected by the output, however there is the important consideration of data ownership and sovereignty, particularly in the Māori context.

Download the table of resources used to inform this summary here. (Last updated: 4/12/2023)