WEF_Shaping_the_Future_of_Learning_2024.pdf

Shaping the Future of Learning: The Role of AI in Education 4.0 INSIGHT REPORT APRIL 2024 Cover: Erik Eastman, Unsplash Contents Executive summary 3 Introduction 4...

Shaping the Future of Learning: The Role of AI in Education 4.0 INSIGHT REPORT APRIL 2024 Cover: Erik Eastman, Unsplash Contents Executive summary 3 Introduction 4 1 Global education systems at a crossroads 6 1.1 Global teacher gap 6 1.2 Administrative and assessment gaps 7 1.3 Digital skills gap 8 2 Potential of AI in enabling Education 4.0 9 2.1 Supporting teachers’ roles through augmentation and automation 9 2.2 Refining assessment and analytics in education 11 2.3 Supporting AI and digital literacy 12 2.4 Personalizing learning content and experiences 13 3 Emerging examples of how AI is advancing Education 4.0 14 3.1 Selection process and criteria 14 3.2 Case studies 15 Conclusion 24 Acknowledgements 25 Endnotes 27 Disclaimer This document is published by the World Economic Forum as a contribution to a project, insight area or interaction. The findings, interpretations and conclusions expressed herein are a result of a collaborative process facilitated and endorsed by the World Economic Forum but whose results do not necessarily represent the views of the World Economic Forum, nor the entirety of its Members, Partners or other stakeholders. © 2024 World Economic Forum. All rights reserved. No part of this publication may be reproduced or transmitted in any form or by any means, including photocopying and recording, or by any information storage and retrieval system. Shaping the Future of Learning: The Role of AI in Education 4.0 2 Executive summary If deployed well, AI can help unlock solutions for improving global education systems. As technological change accelerates, there is an The successful integration of AI into education urgent need for supporting education systems in systems and processes will require careful managing new opportunities and risks. If managed consideration and strategic implementation. The well, technology – particularly artificial intelligence latest in a series of analyses on Education 4.0, this (AI) – offers a unique opportunity to help education paper provides insight into AI’s potential to address systems enable Education 4.0 – teaching and challenges within education systems through: learning approach that focuses on providing learners with the abilities, skills, attitudes and values – Personalized learning content and experiences, fit for the future. Developed by a global coalition offering solutions to the challenge of catering of education experts, practitioners, policy-makers to diverse student needs and enabling tailored and business leaders, Education 4.0 serves as educational journeys for each learner. a comprehensive framework that outlines key transformations needed in primary and secondary – Refined assessment and decision-making education to promote better education outcomes. processes, promising more accurate evaluations AI can help broaden the reach of future-ready and insights into student progress. education systems and enhance their effectiveness in preparing students for the future. Yet, there are – Optimization of teacher roles through challenges and risks, for teachers and learners augmentation and automation of tasks, alike, that must be addressed and overcome to alleviating administrative burdens and deliver on the promise of educational technology. empowering educators to focus more on personalized instruction and mentorship. The adoption of emerging technologies in education, particularly AI, holds immense potential – Integration of AI into educational curricula, to revolutionize teaching methodologies, personalize presenting an opportunity for teaching both with learning experiences and streamline administrative and about AI, equipping students with essential processes. However, while AI can excel at tasks skills, discernment and knowledge for the future. like presenting differentiated content and assuming many administrative duties, the complex process A set of illustrative case studies highlights some of facilitating learning requires more than mere of the learnings thus far in this frontier field. dissemination of information. AI should therefore These examples point to the need for nuanced serve to enhance, not replace, the role of the discussions and further research to explore teacher. By freeing educators from routine tasks, AI opportunities and challenges. By leveraging this empowers them to focus on building relationships, technology judiciously, we can enhance learning understanding individual student needs and outcomes, empower educators and equip fostering motivation. This synergy not only improves students with the requisite skills for success in teaching effectiveness but also underscores the the dynamic landscape of the future. We invite indispensable human element in education. readers to engage with the findings, and support local and global dialogue aimed at shaping a more responsive, inclusive and future-ready education system in the age of AI. Shaping the Future of Learning: The Role of AI in Education 4.0 3 Introduction The latest results of the Organisation for Economic altogether from classrooms and education amid Co-operation and Development (OECD) Programme fears of student cheating and concerns over data for International Student Assessment (PISA) privacy. Others are seeking ways to appropriately saw record drops in student performance on embrace technology in education and cultivate mathematics, reading and science skills,1 even critical thinkers who can understand and work as these skills become more important than ever, alongside AI, bearing in mind changes in the nature particularly in an era of rapid economic, social, of jobs and work in today’s and tomorrow’s labour environmental and technological change. Outcomes markets. According to the World Economic Forum’s for students on critical thinking, collaboration and Future of Jobs Report 2023, employers’ top skill innovation, among other skills in high demand by priorities for 2027 include cognitive skills such as today’s employers, are also mixed across education analytical and creative thinking; technology skills systems around the world. Research suggests such as AI, big data and technological literacy; that, if deployed appropriately, new opportunities and skills required for working with others, such as and developments in artificial intelligence (AI) hold leadership, social influence, empathy and active significant promise for enhancing the effectiveness listening. Additionally, many of the fastest-growing of teachers as well as outcomes for learners, job roles are technology-related roles, necessitating revitalizing education systems towards better digital proficiency.3 preparing students for the demands of the 21st century.2 Education systems must adapt to prepare young people for tomorrow’s technology-driven economies While early forms of AI, such as expert systems and to help students learn alongside these and early machine learning algorithms, have been emerging technologies. The World Economic Forum used in the education field for over 60 years, refers to the teaching and learning of abilities, skills, recent advancements in AI capabilities are creating attitudes and values that are fit for the future as disruption within the education sector. Models such “Education 4.0” (see Figure 1). Developed by a as ChatGPT, Synthesia, Dall-E2 and Bard can write global coalition of education experts, practitioners, essays, create images, explain complex topics policy-makers and business leaders, Education 4.0 and provide step-by-step guidance for solving is a comprehensive framework that outlines key math problems, among many other functionalities. transformations needed in childhood education Generative AI can mimic human logic, writing and to address the needs of the future and promote even creativity, mirroring some human thought better education outcomes. It consists of four sets processes and putting into question the relevance of skills that will be needed in the future – global of some of the skills, principles, formulas and citizenship, innovation and creativity, technology, processes taught in classrooms today, including and interpersonal skills – as well as four sets of basics such as writing, grammar and even logic and learning experiences – personalized and self-paced, discourse. accessible, problem-based and collaborative, and lifelong and student-driven learning. Teaching and The increasing adoption of AI-driven tools by learning that incorporates technology, particularly students for writing assignments and completing AI, can not only help students achieve better assessments has led some educators to question outcomes on technology skills but can also facilitate the basic assumptions that classroom work and enable success in other areas within the accurately reflects students’ cognitive processes. framework. In response, some educators are removing AI Shaping the Future of Learning: The Role of AI in Education 4.0 4 FIGURE 1 Education 4.0 Framework Content (built-in mechanisms for skills adaptation) Experiences (leveraging innovative pedagogies) Global citizenship skills Personalized and self-paced learning To include content that focuses on building From a system where learning is standardized, awareness about the wider world, sustainability to one based on the diverse individual needs of and playing an active role in the global each learner, and flexible enough to enable each community. learner to progress at their own pace. Innovation and creativity skills Accessible and inclusive learning To include content that fosters skills required From a system where learning is confined to for innovation, including complex problem- those with access to school buildings to one in solving, analytical thinking, creativity and which everyone has access in learning and is system-analysis. therefore inclusive. Technology skills Problem-based and collaborative learning To include content that is based on developing From process-based to project and problem-based digital skills, including programming, digital content delivery, requiring peer collaboration and responsibility and the use of technology. more closely mirroring the future of work. Interpersonal skills Lifelong and student-driven learning To include content that focuses on interpersonal From a system where learning and skilling emotional intelligence (i.e. empathy, cooperation, decrease over one’s lifespan to one where everyone negotiation, leadership and social awareness). continuously improves on existing skills and acquires new ones based on their individual needs. Source World Economic Forum. This paper is the first in a World Economic Forum personalized learning experiences, and integrating series on education and AI and sets out to consider AI into educational curricula. The third chapter specific areas where AI may enable Education 4.0 presents case studies that have been collected and showcases practical examples that can serve in collaboration with the Education 4.0 Alliance as inspiration for global leaders and practitioners. to spotlight the intersection of education and AI and offer practical guidance on how new AI The first chapter frames the context for the technologies are beginning to be leveraged in potential use of AI in education by identifying education systems today. The paper concludes key challenges that are contributing to declining with a brief reflection on the investments and policy student performance: the global teacher efforts required to maximize opportunities and gap, gaps in administrative and assessment minimize risks, as well as the next set of activities processes, and the global digital skills gap. The and initiatives that are expected to emerge from the second chapter explores the promise of AI in World Economic Forum’s Education 4.0 Alliance in education – in optimizing teacher roles, supporting 2024-2025. decision-making and management, advancing Shaping the Future of Learning: The Role of AI in Education 4.0 5 1 Global education systems at a crossroads This chapter introduces three key challenges faced tasks, impacting the time they do have to focus by the education sector that may be addressed on quality interactions with students. Third, most through greater integration of technology, including education systems are lagging in closing the digital AI. First, the global shortage of teachers presents skills gap – a critical factor in ensuring the future a significant obstacle to improving education employability of students as well as in developing outcomes and the demand for educators is only in the next generation the necessary aptitude and expected to grow in the upcoming years. Second, ethical awareness for the responsible development teachers spend significant time on administrative and deployment of emerging technologies. 1.1 Global teacher gap The United Nations Educational, Scientific and both decline and growth – in the next five years and Cultural Organization (UNESCO) projects that an on average over 40% of the core skills required in additional 44 million teachers will be needed by all jobs are expected to change during that period. 2030 to fulfil the ambitious targets set forth by As such, the report predicts a growing demand Sustainable Development Goal (SDG) 4, which aims for education-related roles, including Vocational to ensure inclusive and equitable education and Education Teachers, Special Education Teachers, promote lifelong learning opportunities for all.4 This and University and Higher Education Teachers, acute and growing teacher gap is affecting both all among the top 10 positions with the largest developed and developing economies. However, projected increase in employment. the shortage is particularly acute in Sub-Saharan Africa, where an additional 15 million teachers will Education systems compete with various sectors be needed to provide universal childhood education of the economy to attract top-tier graduates by 2030.5 into teaching roles. Studies show that offering competitive salaries is crucial for both retaining This need is set to grow as global labour markets teachers and attracting new individuals to the face disruption, and the need for reskilling, upskilling profession. Yet, in many OECD countries teaching and lifelong education grows. According to the is not a financially attractive career choice. On World Economic Forum’s Future of Jobs Report average, lower secondary (typically, the first three 2023 one-quarter of all jobs face transformation – years immediately following primary education Shaping the Future of Learning: The Role of AI in Education 4.0 6  and which in many countries ends compulsory ready” profession. While technology will never fully education) teachers’ salaries lag behind those replace human teachers, AI and other emerging of tertiary-education workers by 10%; in some technologies can immediately address some of countries, the gap is over 30%.6,7 this gap. Many teachers already acknowledge the benefits of such support. For instance, in the United A significant set of work will need to be done by Kingdom, 42% of primary and secondary teachers governments and other stakeholders to ensure used generative AI to aid with their schoolwork that a robust set of new talent joins the future in November 2023, a significant increase from teaching workforce, that teachers are adequately 17% in April 2023.8 Alongside new incentives remunerated, and that teaching is positioned as and structural frameworks aimed at developing, a high-growth, high-potential job of the future. attracting and retaining talent within the education There is an opportunity for AI and other emerging sector, governments, business and civil society can technologies to help address these goals, by support the integration of AI as a tool for today’s supporting those already in the teaching workforce teachers and as an attractive additional skill set for and ensuring that teaching emerges as a “future- prospective future teachers. 1.2 Administrative and assessment gaps The scarcity of teachers is compounded by the Inefficient assessment processes also hinder administrative burdens they face in the workforce. the ability of education leaders at the district, A recent survey of teachers in the United States national and global levels to make timely and found that while they work an average of 54 hours data-driven decisions when it comes to their per week, only 46% of that time is spent teaching. education strategies and investments. For example, Similarly, when looking across OECD countries, while education systems aim to regularly assess lower secondary school teachers spend an average student understanding throughout the school year, of about 44% of their working time on teaching and comprehensive evaluations that are comparable the rest of their time on non-teaching tasks.9 The across schools occur infrequently. Learner burden of repetitive administrative tasks is regularly performance is typically evaluated sporadically by cited by teachers and school leaders as one of schools and only annually by Ministries of Education the leading aspects affecting the quality of jobs in during curriculum review. Meanwhile, cross-country the education sector. In the latest OCED Teaching comparisons, such as those conducted through and Learning International Survey, the main source the OECD’s Programme for International Student of stress for teachers in both primary and lower Assessment (PISA), take place every three years. secondary education has been “having too much This lack of frequent data on student learning administrative work to do”, at 47% and 48%, outcomes and skills gaps prevents education respectively.10 systems from having the agility required to adapt to changing learning and labour-market needs. Addressing these pain points expressed by teachers and school leaders by automating Integrating AI technologies into educational administrative tasks and augmenting human- assessments offers the potential for educators to centric ones could help free up time that teachers gain real-time, data-driven insights into student can spend on higher value tasks – such as directly learning trends, identifying areas of strength and engaging with students, customizing content weakness and assessing instructional effectiveness for maximum impact or developing their own on a large scale. It also helps in evaluating pedagogical skills. Alleviating this burden can non-standard tests more efficiently, informing improve the attractiveness of the sector, which instructional decision-making and curriculum could in turn help mitigate the global teacher development, and enhancing the overall quality of shortage noted in the previous section. education delivery. Shaping the Future of Learning: The Role of AI in Education 4.0 7 1.3 Digital skills gaps Generative AI has the potential to create trillions of However, AI can also lead to job displacement by dollars in economic value, driven by increases in automating the majority of tasks in some roles. labour productivity and the creation of new revenue Those lacking technological literacy are most at risk streams from product innovation.11 However, such of displacement, while those up are able to reskill estimates rely on the assumption that individuals, and upskill in a variety of skills – including becoming teams and organizations will have the ability and proficient in using, developing, explaining or willingness to use AI and other technology tools applying AI – are most likely to make successful job effectively. transitions.14 To prepare workers and address global digital and skills shortages in the medium-to-long While the potential economic benefits of generative term, teaching about technology, including AI, must AI and other emerging technologies are promising, be emphasized in education. unlocking this value hinges on addressing the most significant barriers: persistent global shortages in These future-ready digital skills should also include digital skills and AI talent. The current labour market education about the use of new technologies as is already experiencing significant shortages of well as how to be safe and ethical producers and workers with expertise in AI technology and this consumers of technology. Young developers need demand is poised to escalate further. A recent to understand the ethical considerations when survey shows that 68% of executives report a designing AI and must develop a keen awareness moderate-to-extreme AI skills gap.12 Furthermore, of the potential risks and implications of AI design advances in AI development are slowing due to a and deployment. global shortage of talent with skills in areas such as deep learning, natural language processing and robotic process automation.13 Shaping the Future of Learning: The Role of AI in Education 4.0 8  2 Potential of AI in enabling Education 4.0 This chapter explores the potential ways in which First, integrating AI into education presents AI can address the gaps highlighted in the previous an opportunity to streamline a broad set of chapter. administrative tasks for teachers, allowing them to dedicate more time to engage with students. Second, AI can help teachers in assessing learners FIGURE 2 4 Promises of AI in education more rapidly and enabling them to provide more immediate feedback. Third, AI can enable students and learners to develop digital literacy, critical Supporting teachers' role: thinking, problem-solving and creativity skills. augmentation and automation Finally, AI can personalize the learning experience, supported by teachers, leading to improved Refining assessment and academic performance and better adaptation to decision-making in education diverse learning needs. Across all four opportunity areas, AI is a complementary tool that enhances Supporting AI and the educational experience, while preserving the digital literacy essential human elements embedded in teaching and learning. In addition, learning about AI and Personalizing learning digital skills – even through traditional methods – content and experience can support learners in being prepared for the jobs of tomorrow. 2.1 Supporting teachers’ roles through augmentation and automation New developments in AI can provide an opportunity Tasks with the most potential to be automated or to redefine the nature and quality of work in replaced by LLMs are those that tend to be routine education roles. Research by the World Economic or repetitive. In the education sector, up to 20% of Forum, produced in collaboration with Accenture,15 work time on clerical activities and administrative finds that 40% of all time spent on tasks could tasks, such as assessing attendance, enrolment and potentially be impacted by large language models other forms of data analysis, could be automated. (LLMs). This applies to teaching as well: while some Tasks most likely to benefit from the augmentation teaching tasks could potentially be automated by potential of LLMs tend to emphasize analytical these new technologies, other tasks stand to be and problem-solving capacities. These tasks make augmented or enhanced by LLMs (see Table 1). up 8%-20% of work time spent on tasks in the education sector and include lesson planning and evaluating student performance. Shaping the Future of Learning: The Role of AI in Education 4.0 9  TA B L E 1 Impact of large language models (LLMs) on education tasks Automatable tasks – Compile lists of books, periodicals, articles and audio-visual materials on particular subjects. – Verify facts, dates and statistics, using standard reference sources. – Grade homework and tests, and compute and record results, using answer sheets or electronic marking devices. Augmentable tasks – Analyse performance data to determine effectiveness of instructional systems, courses or instructional materials. – Design learning products, including Web-based aids or electronic performance support systems. – Develop teaching or training materials, such as handouts, study materials or quizzes. – Prepare assignments for teacher assistants or volunteers. Lower potential – Establish clear objectives for all lessons, units and projects and communicate those objectives to for exposure and children. unaffected tasks – Confer with leaders of government and community groups to coordinate student training or to find opportunities for students to fulfil curriculum requirements. – Collaborate with other teachers and administrators in the development, evaluation and revision of elementary school programmes. – Plan and supervise class projects, field trips, visits by guest speakers or other experiential activities, and guide students in learning from those activities. – Set up classrooms, facilities, educational materials or equipment. Source World Economic Forum, in collaboration with Accenture, 2023. Tasks that emphasize interpersonal interactions, like as curriculum design and, of course, the essential face-to-face communication or physical interactions educational aspects of interpersonal interaction. with young learners, are likely to be unaffected or not However, such a transformation would need to enabled by LLMs, and most teaching tasks – and be carefully designed and enabled to ensure roles – feature prominently in this category of jobs that teachers are able to manage the pace of that are likely to be unaffected by AI (see Figure 3). automation or augmentation and be supported in their own upskilling, while they learn to focus on the Taken together, the automation and augmentation more human-centric aspects of their jobs, such as potential of LLMs for routine and repetitive refining their pedagogy, providing social-emotional administrative work in teaching opens up more support, individualized instruction and parent time for educators to focus on creative tasks such engagement. Shaping the Future of Learning: The Role of AI in Education 4.0 10  FIGURE 3 Potential for automation and augmentation of education jobs Special Education Teachers, Middle School Tutors Special Education Teachers, Secondary School Preschool Teachers, Except Special Education Career/Technical Education Teachers, Middle School Career/Technical Education Teachers, Secondary School Secondary School Teachers, Except Special and Career/Technical Education Kindergarten Teachers, Except Special Education Elementary School Teachers, Except Special Education Special Education Teachers, Kindergarten and Elementary School Middle School Teachers, Except Special and Career/Technical Education Special Education Teachers, Preschool Special Education Teachers, All Other Substitute Teachers, Short-Term 0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100% Higher potential for automation Higher potential for augmentation Lower potential for automation or augmentation Non-language Source World Economic Forum, in collaboration with Accenture, 2023. 2.2 Refining assessment and analytics in education Today’s models of standardized and informal both teachers and students by eliminating the need assessment often exhibit linear and time- to conduct one-off, high-stakes exams. Through consuming characteristics, as discussed in the automated, regular feedback mechanisms, students first chapter. Similar to how human tutors can offer can engage in meaningful, enjoyable learning instant, personalized feedback, AI automation in activities where all learning is analysed in real-time, assessments can allow for immediate feedback on rather than relying on periodic formal assessments. a larger scale, aiding students in comprehending This shift from traditional assessment methods to mistakes and supporting teachers in identifying dynamic, real-time analytics has the potential to areas for improvement. significantly enhance the educational experience, fostering adaptive learning environments that cater However, such analytics are best enabled in to the diverse needs of students. partnership with teachers. AI tools can be programmed with the support of teachers who can Machine learning and AI-powered analytics provide examples of feedback for AI to learn from, can enable education systems to be more agile including in evaluating non-test assignments such and responsive to immediate learner needs. All as essays, project proposals and similar tasks. stakeholders – including students, teachers, parents, school leaders and ministries – can receive Furthermore, the adoption of game-based timely analytics for informed and adaptive decision- assessment technologies can alleviate pressure on making, fundamentally transforming current Shaping the Future of Learning: The Role of AI in Education 4.0 11  linear and time-lagged approaches to learning for understanding larger patterns within and across assessment into future-ready, responsive and education systems as well as predicting where dynamic models. Big data sets can be analysed future gaps may emerge across cities and regions. not only for correct or incorrect answers, but also 2.3 Supporting AI and digital literacy Developing digital skills is essential for navigating representing about 49% of the world’s population.17 today’s technological landscape and lays the Teaching about AI not only equips students groundwork for both AI and digital literacy. Digital with the ability to recognize disinformation and and AI literacy goes beyond the mere ability to use misinformation but also fosters their development digital tools and platforms; it also encompasses into responsible future AI developers. Moreover, critical thinking, problem-solving, creativity and incorporating basic cyber skills into curriculum can awareness of the ethical implications of AI. help students learn how to build robust and safe AI systems. Safeguarding the security and integrity of Integrating AI into education presents an opportunity AI data systems is imperative, particularly in light of to not only utilize AI tools in teaching but also to the potential risks associated with data breaches, educate students about AI concepts and their hacking and malicious manipulation of AI algorithms. broader societal impacts. Integrating AI into curricula does not imply that every student must become Enhancing the integration of AI applications into an AI expert. Rather, the emphasis should be education systems globally can give the technology on cultivating awareness, nurturing curiosity and a pivotal role in educating students about establishing a foundational understanding – for responsible and equitable AI practices. Resources example, by teaching students how to assess the already exist for teaching about AI, such as the reliability of sources and discern the accuracy of International Society for Technology in Education information presented on websites. One study (ISTE) “Hands-on AI Projects for the Classroom”, found that digital literacy is indeed a good predictor which includes specific projects that teach of one’s ability to distinguish between facts and concepts such as unconscious bias and active vs. misinformation.16 This is a particularly important passive data collection, and terms such as machine and urgent life skill as more people than ever in learning algorithm and targeted marketing.18 history will vote in 64 national elections in 2024 – Shaping the Future of Learning: The Role of AI in Education 4.0 12  Some economies have started to lay out basic process data.19 Australia, Japan and New Zealand principles for how to integrate AI and digital literacy have also outlined guidance around teaching with into classrooms. In the United Kingdom, for and about AI. example, the Office for AI is currently conducting research to support primary and secondary schools Encouraging digital and AI literacy among learners to teach critical skills such as the limitations, equips them with valuable skills for navigating the reliability and potential bias of generative AI; how increasingly AI-driven aspects of tomorrow’s job information on the internet is organized and ranked; market, providing them with a competitive edge and and foundational knowledge about how computers greater versatility in their career paths. work, connect with each other, follow rules and 2.4 Personalizing learning content and experiences A study by education psychologist Benjamin feedback, much in the same way that a private tutor Bloom found that the combination of one-to-one would. tutoring alongside regular tests and feedback led to student performance that was two standard Algorithms can not only customize the content but deviations – about 98% – above those of students also adjust the pace, difficulty and learning style who receive standard classroom instruction.20 The depending on learner performance, behaviour and research concluded that “there is a great difference preferences.25 Based on data patterns, AI can in student cognitive achievements, attitudes, and predict learning challenges, identify gaps and create academic self-concept under individual tutoring personalized learning journeys by analysing trend compared with the group method of instruction”. data and students’ learning history, preferences and Providing personal tutoring dramatically changed performance. AI can provide materials that match the distribution of education achievements in the students’ strengths, weaknesses and knowledge class. A more recent study by Stanford University levels, and align with learning objectives, thereby researchers found that even short tutoring enhancing the relevance of the educational content interventions, as brief as 10 minutes a day, result in for each individual learner. significant improvement in young student’s literacy skills.21 However, these new tools are best enabled when complemented by rigorous stress-testing However, scaling personal tutoring methodology processes by teachers in personalizing support, is costly and inefficient in even the most advanced tailoring culturally relevant teaching and learning economies. It would require a significant change materials, and providing instant translation to adapt in teacher-pupil ratios, which is challenging if not content to learner needs. The relevance of the unrealistic, given existing global teacher shortages. materials and examples is essential for creating an While private tutoring is well recognized for its engaging, understandable and applicable learning impact on student performance, and the global environment for learners – and AI tools, together private tutoring market is projected to grow from with teachers, can relate examples and concepts $57.92 billion in 2023 to $105.98 billion by 2030, to each student’s interests, lived experiences and access is generally restricted to those who can backgrounds. afford it, further perpetuating inequalities in learning outcomes.22 Finally, AI can also present materials in various modalities to address different visual, audial and Since the rise of personal computing and physical needs. Customizable interfaces and digitalization, there has been growing interest adaptive technologies are particularly valuable in using technology to accelerate personalized for neurodiverse students and those with varying learning.23 A study conducted between 2007 and physical abilities. For example, through AI 2020 found that technology-supported personalized technology, classroom lessons can be captioned learning had a significant positive effect on learning for students who have auditory impairments, outcomes.24 While technology has thus far not allowing them access to any classroom rather than been able to fully replicate the benefits of one- relying on the availability of human sign language to-one tutoring, recent advances in AI are able to assistants; this helps teachers and learners engage analyse and learn from big data sets, providing in faster and more personalized communication. tailored learning content, experiences and real-time Shaping the Future of Learning: The Role of AI in Education 4.0 13 3 Emerging examples of how AI is advancing Education 4.0 The integration of the latest AI tools in education a comprehensive overview of all innovative is a relatively new development, which needs and impactful examples, they showcase the careful management and monitoring of results. transformative potential of AI-driven innovations and The case studies presented in this chapter offer inspire educators, policy-makers and businesses to practical guidance on how AI technologies are embrace the opportunities and calibrate risks and being leveraged in education systems today. challenges. While the selected case studies do not represent 3.1 Selection process and criteria The World Economic Forum’s Education 4.0 – Scalability: future impact potential beyond Alliance – made up of leaders, experts and the current reach and applicability in different decision-makers in education – aims to identify contexts policies, initiatives and programmes that advance the Education 4.0 framework. In 2023, the Alliance – Sustainability: sustainability of the initiative and developed a set of criteria to identify emerging potential for long-term impact examples of how AI tools are being leveraged to advance Education 4.0 in the absence of global Recognizing the potential that AI may have to standards and comprehensive policies. These exacerbate current education gaps, all selected case criteria include: studies strongly emphasize education equity in their design. Following a call for submissions through – Significance: magnitude, reach and its various partner networks, the World Economic transformative nature of the impact Forum’s Centre for the New Economy and Society and the Alliance, with the assistance of a panel of – Quantifiability: use of metrics to measure and independent experts, selected nine case studies drive further impact relating to the gaps and opportunities identified earlier in this paper and based on the evaluation criteria. Shaping the Future of Learning: The Role of AI in Education 4.0 14 3.2 Case studies CASE STUDY 1 Accessible Digital Textbooks (ADT) UNICEF The initiative employs Universal Design for Learning (UDL) principles and accessible technology to create digital tools that cater to diverse learners, including those with disabilities. Context and objectives Of the 240 million children globally who have disabilities, the vast majority lack access to inclusive technologies, easily accessible learning materials and other vital educational support to fully engage and participate in their learning endeavours. Half of children with a disability are out of school and one billion children and adults with disabilities need assistive technology but do not have access to it. Managed by UNICEF, the Accessible Digital Textbooks (ADT) initiative employs UDL principles and accessible technology to create customizable, inclusive digital tools for diverse learners, including those with disabilities. UNICEF plans to leverage AI for cost-effective scaling, collaborating with partners globally to transform education for children with disabilities, and expanding the initiative to new regions. A co-creation approach involving local stakeholders enhances the ADT ecosystem, contributing to improved learning outcomes and reimagining the future of textbooks. ADTs have been implemented in three countries in the Eastern and Southern Africa regions, and six countries in Latin America and the Caribbean, in coordination with the respective Ministries of Education, with a goal to reach 500,000 children in the first half of 2024. AI-enabled aspects UNICEF aims to revolutionize textbooks using AI for widespread implementation. ADTs allow users to customize and combine diverse features like narration, sign-language videos, interactivity, the audio description of images, text-to-speech and other functions to suit different preferences or access needs. Once installed, the learner can use the textbook offline on the device, making it accessible to students who lack connectivity, promoting education that is personalized, inclusive and accessible Expected impact Research and development conducted by UNICEF and its partners indicate that ADTs can enhance students’ motivation, classroom participation and their ability to engage with one another. Shaping the Future of Learning: The Role of AI in Education 4.0 15 3.2 Case studies CASE STUDY 2 Skill-building with Virtual Mentors Kabakoo Academies An educational technology start-up in West Africa pioneering a transformative approach to upskilling young people in the face of limited formal job opportunities. Context and objectives Over 80% of employment in Africa is informal, impacting the opportunities available to the continent’s young talent. Kabakoo is an educational technology start-up with a mission to enable young people in West Africa to develop the mindset and skills for self-employment in a setting that lacks formal jobs. They have developed a community-driven upskilling approach that combines a mobile application with real-life networks of peers and mentors. Kabakoo leverages social media content and local partnerships to engage youth in urban and semi-urban areas of West Africa. The Kabakoo app provides community- based experiential learning, enriched by modules on learning to learn and visualization and with the support of an LLM-based virtual mentor. Furthermore, Kabakoo fosters real-life community interactions to support the acquisition of digital and entrepreneurial skills AI-enabled aspects Kabakoo employs an AI-enabled virtual mentor to provide 24/7 support to learners. This virtual mentor offers guidance, resources and advice whenever needed, supplementing human mentorship. The AI mentor also provides personalized feedback on learners’ assignments. After submitting their selfie video on a specific module, learners receive a personalized response via WhatsApp. Recognizing the linguistic diversity in Mali, Kabakoo is working on developing an AI-powered model to provide training in Bambara, the most spoken language in the country. In applying AI to address language barriers, Kabakoo promotes personalization, accessibility and inclusivity. The use of gamified virtual tokens (Kabakooins) and cloud-based resources contributes to a dynamic and robust learning environment. Expected impact The success of the programme is evidenced by a randomized control trial that resulted in a 23% increase in growth mindset among learners in a pilot conducted at Kabakoo. Kabakoo learners also report seeing a 44% increase in income six months after completing the programme. Shaping the Future of Learning: The Role of AI in Education 4.0 16 3.2 Case studies CASE STUDY 3 Leveraging Literacy through AI Letrus An AI-based literacy development initiative implemented in middle and high schools across Brazil. Context and objectives The latest PISA scores show performance in Brazil and Latin America below the OECD average. The Letrus Program is an AI-based literacy development initiative implemented in middle and high schools across Brazil and currently benefiting 170,000 students in 670 schools. The main objective of the programme is to narrow the literacy gap between low- and high-income students. It incorporates proprietary natural language processing AI technology to offer real-time constructive feedback in reading and writing. AI-enabled aspects Letrus focuses on personalized learning through AI, offering immediate feedback to students, real-time data for educators and monitoring tools for school managers. Teachers receive tailored recommendations for content and methodologies, which can be seamlessly integrated into the curriculum to nurture and enhance specific skills. This iterative process ensures a dynamic and responsive approach to literacy development, aligning closely with the evolving needs of each student as well as the entire class. School managers can monitor progress and gain immediate insights into improvement areas as well as emerging learning gaps that may benefit from targeted intervention through teacher-training initiatives or strategic adjustments to the curriculum. Expected impact In 2022 the programme achieved notable success in the public schools of Espirito Santo. Within five months of programme implementation, participating students achieved the second position in the national writing exams, a remarkable improvement compared to the eighth position attained by the control group. Letrus was subsequently designated as the official literacy development programme for high school students in the state. Espírito Santo emerged as the top-performing state in the writing component of the National Exam, exhibiting a performance delta five times the national average from 2021 to 2022. Shaping the Future of Learning: The Role of AI in Education 4.0 17 3.2 Case studies CASE STUDY 4 Pensamiento Computacional e IA (Computational Thinking and AI) Ceibal This programme teaches computational thinking and AI in an interdisciplinary way with other areas of knowledge such as mathematics, language and science. Context and objectives Computational thinking is a mental process by which humans try to solve complex problems by breaking them down into smaller, more manageable parts in the way computers would. It is a foundational skill for AI development. The Ceibal Computational Thinking and Intelligence programme operates in 80% of urban public schools and 250 rural schools in Uruguay. The programme’s key objective is to teach computational thinking and AI in an interdisciplinary way with other areas of knowledge such as mathematics, language and science. It also includes an active intervention to reduce the gender gap in these skills. The programme focuses on teaching students to be ethical producers and knowledgeable consumers of AI, covering topics such as how a machine learning model works, how data is used and analysed, and the biases that may exist. AI-enabled aspects AI lessons are utilized to help students understand the inner workings of machine learning models and gain insights into the use of data, as well as the potential biases present. The programme employs various evaluation tools, including learning tests, surveys and class observations. The programme’s aim is to foster the competencies of solving computational problems, data and information analysis, algorithms and procedures, as well as social transformation, acknowledging the integration of computers into our everyday lives. Expected impact Ceibal participates in the Bebras Competition, an international exam on computational thinking. A recent sample of Bebras exam results showed that students who participated in the programme significantly outperformed those who did not, with some differences observed in favour of girls. This community- driven initiative led to the integration of computational thinking competence into Uruguay’s 2023 educational reforms. Shaping the Future of Learning: The Role of AI in Education 4.0 18 3.2 Case studies CASE STUDY 5 The School Cyber Security Challenges / Cyber Skills Aotearoa Grok Academy An initiative aimed at providing resources for teachers to teach cybersecurity concepts and inform students about career opportunities in the field. Context and objectives Grok Academy is a not-for-profit Australian organization that supports the teaching of computing science and related disciplines. The organization has an extensive track record of creating student-centred and curriculum-aligned materials for cybersecurity and digital technology targeted to primary and secondary school students and teachers. In 2019, Grok Academy launched the Australian Schools Cyber Security Challenges programme, an initiative aimed at providing resources for teachers to teach cybersecurity concepts and inform students about career opportunities in the field. Developed in collaboration with government and industry organizations, the programme includes course curricula and classroom activities targeting students in years 5-12. Its success in Australia prompted expansion into New Zealand, leading to the introduction of Cyber Skills Aotearoa in October 2022. With courses, competitions and unplugged resources (ones that don’t include digital devices), it engages over 91,000 students. Grok Academy’s strategy emphasizes industry and government collaboration, reaching thousands across Australia and New Zealand. AI-enabled aspects As generative AI – which is based on large amounts of data – continues to develop, cybersecurity skills will become increasingly important. By teaching about cybersecurity, which encompasses information privacy and security, cryptography and digital forensics, Grok Academy is preparing the next generation of talent to be responsible producers of AI. There is also a clear understanding of the benefits of the programme in terms of educating the broader family community of participating students, since for many families this will be their only exposure to cybersecurity and misinformation concepts. Expected impact Over 4.5 million students have been introduced to cybersecurity skills since the launch of Grok Academy. While roughly half of all participants are girls, female participation exceeds male participation in years 4, 7 and 8, helping to close persisting gender gaps in cybersecurity. Shaping the Future of Learning: The Role of AI in Education 4.0 19 3.2 Case studies CASE STUDY 6 3D Africa for Girls Youth for Technology Foundation A unique programme that encourages girls aged 10-16 years to develop innovative STEM-based solutions to real-world problems. Context and objectives Digital manufacturing and 3D printing have the potential to revolutionize Africa’s manufacturing industry in the same way that smartphones and mobile broadband are transforming the service, trade and agricultural industries. 3D Africa for Girls aims to transform the continent from “Aid to Africa” to “Made in Africa”. It provides high-quality STEM education in Nigeria that enables young girls ages 10-18 to design, prototype, market and sell their 3D-printed products and solutions. By doing so, it teaches and models girls to develop innovative STEM-based solutions to real-world problems, and teaches marginalized, low-income youth, girls and women how to sell those products in global, online markets. The programme leverages a combination of targeted job-shadowing and mentoring, and 3D Design and prototyping. AI-enabled aspects The programme’s unique approach integrates 3D printing skills, real-world experts, mentors and global online marketplaces. Girls express their creativity using computer-aided design (CAD) to turn their ideas into new products (such as 3D printed rechargeable, detachable, cordless hair dryers and African-designed furniture) and market their goods. In addition to 3D design and printing, the programme also emphasizes programming skills, which are the backbone of any career in AI. Recognizing the persistent gender gap in AI careers (only 30% of the global AI workforce is female), the programme introduces young girls to the fundamentals of programming through tools such as Scratch and Bootstrap. Expected impact The programme has been instrumental in addressing the lack of encouragement for girls’ interest in STEM subjects, with at least 90% of graduates enrolling in university or tertiary institutions one year after completion and over 85% continuing in STEM or technology careers afterwards. Within the programme, at least 90% of participants attain a minimum of 95% proficiency in basic and advanced technology skills. Shaping the Future of Learning: The Role of AI in Education 4.0 20 3.2 Case studies CASE STUDY 7 AI for Youth Entrepreneurship Curriculum JA Europe An innovative curriculum blending AI and entrepreneurship education for youth. Context and objectives The curriculum enables youth to develop a foundational understanding of AI, including ethics, data literacy and operations. Equipped with this foundational knowledge and practical skills, participants are poised to devise economically viable solutions – leveraging AI and associated tools – that address local and global challenges. Over the long term, this curriculum aims to support a future in which all youth are AI natives. While the programme has just completed its pilot phase, JA Europe recently expanded the curriculum to 10 additional countries, which will enable it to reach 30,000 youth in the next two years, combining the application of AI with improvements to entrepreneurship and employment in the agricultural sector. AI-enabled aspects The curriculum focuses on preparing educators and students for a technology- driven economy, offering coding-focused and non-coding pathways to develop AI understanding, ethics, data literacy and operations. It measures efficacy through entrepreneurship competencies and tech-driven solutions. Participants engage with advanced digital skills and tools, facilitating online collaboration, ideation and the application of AI, coding and other essential tech tools in their entrepreneurial endeavours. The curriculum was developed in collaboration with dedicated Intel experts and features an interactive online computing platform for youth to develop AI-based solutions and create business plans. To ensure educational equity, JA’s strategic focus lies in supporting underserved schools, especially in remote areas, and offering training to refugees, particularly from Ukraine. AI for Youth Entrepreneurship is implemented in schools selected based on their limited resources and geographical location, and in alignment with national strategies aimed at reaching underserved youth. Expected impact In the pilot year, student teams created 34 AI-based solutions, demonstrating the success of the programme in developing entrepreneurial capacities and understanding of AI among young people. Shaping the Future of Learning: The Role of AI in Education 4.0 21 3.2 Case studies CASE STUDY 8 AI-Powered Digital Textbooks Ministry of Education of South Korea A digital textbook that enables customization for learners across various proficiency levels. Context and objectives The Ministry of Education in South Korea has unveiled plans to introduce AI-powered digital textbooks in local elementary and secondary schools starting in 2025. The initiative aims to meet the growing demand for diverse learning content and utilize AI and other emerging technologies to enhance the educational experience for students. This innovation in South Korea’s education system aims to address educational inequality, reduce reliance on private education and transform the nation’s hypercompetitive education culture. The Ministry of Education plans to continue refining the initiative, incorporating feedback and ensuring the development of diverse and creative AI digital textbooks. AI-enabled aspects The introduction of digital textbooks will enable customized learning opportunities for students across various proficiency levels in subjects such as mathematics, English and informatics. The programme will progressively expand to include additional grades and subjects, and eventually cover all subjects by 2028, excluding activity-based subjects like music, art, physical education and ethics. Students will receive tasks and activities differentiated based on their individual proficiency, allowing students to learn at their own pace. The Ministry of Education emphasizes the coexistence of paper and digital formats in schools. The plan also emphasizes collaboration between human teachers and AI assistants, with a focus on fostering lead teachers who provide humanized guidance alongside AI technology. Expected impact The Ministry of Education is currently setting up a pilot programme involving 400 teachers who will be using the AI-powered textbooks in their teacher training. The rollout of the programme is expected to begin in 2025 and will enable teachers to tailor their classes through real-time collection and analysis of data about students learning. Shaping the Future of Learning: The Role of AI in Education 4.0 22 3.2 Case studies CASE STUDY 9 AI Tutor Project Ministry of Education of the United Arab Emirates An AI-powered virtual tutor to enhance the education landscape in the UAE and promote educational equity. Context and objectives The Ministry of Education of the United Arab Emirates – in collaboration with several partners, including Microsoft, ASI and teachers, among others – has developed an AI-powered tutor to enhance the education landscape in the UAE and promote equity in a context where private tutoring is on the rise. The project aims to significantly improve students’ academic performance, resulting in higher test scores, better comprehension of subjects, and enhanced critical thinking skills through AI-powered personalized learning. The project employs adaptive learning algorithms, continuous assessment, 24/7 availability and data-driven insights to achieve its goals. It aims to create a more engaging and interactive learning experience, promoting self-directed learning and eliminating barriers related to time and location. While the project is in its infancy and pilot stage (the first version is scheduled to launch September 2024), the intended evidence- gathering methods for skill development include ongoing data collection, assessments and evaluations on personalized learning, critical thinking, problem- solving, self-directed learning, digital literacy and communication skills. AI-enabled aspects The AI tutor tailors lessons to the individual needs and learning styles of each student, ensuring that they receive the right level of challenge and support. It continually assesses student progress, identifying areas of improvement and providing targeted feedback and additional resources to help students overcome their challenges. The tutor can also provide support in multiple languages, ensuring that students from diverse linguistic backgrounds also have access to quality education. By automating certain teaching tasks and providing valuable insights into student progress, the UAE AI Tutor project aims to alleviate teacher workload, enabling them to focus on more strategic and interactive aspects of the learning experience. The platform also generates detailed reports for educators and parents, enabling them to monitor and support student progress effectively. Further, the AI tutor breaks up linear and time-consuming methods of feedback to provide real-time analytics to all stakeholders, including at the ministerial level, enabling more adaptive strategy development. Expected impact The project already showed improvements during the piloting stage in average grades and positive impact on students’ academic performance. It demonstrated a 10% increase in learning outcomes. Shaping the Future of Learning: The Role of AI in Education 4.0 23 Conclusion The integration of AI technology into education between human expertise and AI capabilities presents a promising pathway for enhancing that holds the potential to transform learning learning experiences and outcomes, while scaling outcomes. The case studies presented in this report AI literacy can support learners in being prepared underscore this element of integration as well as the for the jobs of tomorrow. At the same time, it is need for comprehensive public-private partnerships, important to acknowledge the potential risks of planning and impact assessment. Under these rapid generative AI deployment in education without conditions, AI technologies can enhance teaching putting appropriate planning, safety measures, and learning experiences, promote equitable governance measures and equity frameworks in access to education and address pressing place. While AI systems often outperform similar, challenges in the educational landscape. traditional software systems that are commonly viewed as “educational technology” or “edtech”, The case studies point to five conditions that can they have attributes that may both amplify and help balance innovation with guardrails: create new risks. – AI for education must be developed in Some of these concerns are widely held by parents, collaboration with educators and education educators and leaders alike. A global survey of leaders, ensuring it meets the diverse learning more than 17,000 people in 2023 found that 61% needs of students, supports teachers in of respondents are either ambivalent or unwilling their focus on human interactions and aids to trust AI systems, while 71% of people are administrators in making informed decisions on concerned about potential risks.26 For example, relevant content. while the data collected by these tools enables their powerful results, it also leads to concerns – AI integration in education must ensure the about safeguarding student data and privacy. protection of sensitive information through These concerns are linked to wider challenges the implementation of robust data privacy around safeguards for AI. For example, a survey and security protocols. This includes raising of over 2,300 IT professionals found that only 10% awareness about ethical data use, ensuring of firms had policies intended to govern the use consent, anonymizing data and limiting data of generative AI, while 57% are very or extremely collection to what is strictly necessary for worried about generative AI being exploited by bad educational purposes. actors.27 Some concerns are held by educators themselves, around the potential of AI to disrupt – Innovative funding models are essential to teaching jobs, requiring careful management, support the ongoing refinement of AI algorithms ensuring teachers remain central to education and the creation of independent testing and systems complemented by AI tools, and adequate evaluation of AI solutions so that stakeholders reskilling and upskilling for teachers. can be confident in their efficacy and value for money. Conversely, there are concerns around the potential for exacerbating inequality for those that will miss – Students, teachers and administrators must out on the benefits of AI in education. Over 2.6 receive necessary training and upskilling billion individuals globally currently lack basic opportunities oriented to their needs to help internet access.28 It is essential, therefore, that them make the most productive use of AI effort and focus be oriented toward the issue of systems. educational equity, lest the benefits of AI systems accrue primarily to geographies and communities – Equity and inclusion considerations must be which are already relatively privileged, further central to the design of programmes to ensure widening the education equity gap. Additionally, that AI literacy is widely imparted and the there are other concerns about inaction on the benefits of AI technologies in education accrue part of education systems in integrating a focus widely. on AI, as synthetic content can be used to create misinformation and disinformation.29 Children and The next phase of research in Education 4.0 will young learners are particularly vulnerable – and a explore these aspects and continue to develop case focus on AI literacy is necessary to help prepare studies and examples of successful integration of them to critically analyse content and understand AI tools and AI-literacy in education, working with the adverse consequences of false information. business, government, educators and civil society to unlock the transformative potential of AI in shaping While education systems need to adapt to use the future of education for generations to come. AI and to teach AI literacy, it is the collaboration Shaping the Future of Learning: The Role of AI in Education 4.0 24 Acknowledgements The World Economic Forum would like to thank the members of the Education 4.0 Alliance as well as the members of the broader core community of the Centre for the New Economy and Society for their ongoing commitment and contributions to addressing the issues presented in this report. We would especially like to thank Andria Zafirakou, Teacher at Alperton Community School, and Christopher Dede, Senior Research Fellow at the Harvard Graduate School of Education, for their thorough review of case studies and expertise. We would also like to thank the team at Accenture, and Nicole D’Agostino and Christine Nanan for their collaboration on data from the Jobs of Tomorrow reports. This work was made possible through the generous financial support of the Ministry of Education of the United Arab Emirates to the World Economic Forum’s Reskilling Revolution Initiative. We gratefully acknowledge the support of Mike Fisher for copyediting and Accurat for graphic design and layout. We are further grateful to our colleagues in the Centre for the New Economy and Society for their helpful suggestions and comments, with a special thank you to Attilio Di Battista, Eoin Ó Cathasaigh, Sam Grayling, Andrew Silva, Mirielle Eaton and Fernando Alonso Pérez-Chao for their unique support. The views expressed in this White Paper do not necessarily represent the views of the World Economic Forum nor those of its Members and Partners. This briefing is a contribution to the World Economic Forum’s insight and interaction activities and is published to elicit comments and further debate. Contributors Genesis Elhussein Manager, Reskilling Revolution and Skills Initiatives, Centre for the New Economy and Society Elselot Hasselaar Head, Mission for Work, Wages and Job Creation, Centre for New Economy and Society Ostap Lutsyshyn Specialist, Education 4.0, Centre for the New Economy and Society Tanya Milberg Manager, Education Initiatives, Centre for the New Economy and Society Saadia Zahidi Managing Director, World Economic Forum Shaping the Future of Learning: The Role of AI in Education 4.0 25 Education 4.0 Alliance Jonghwan Patrick Park Founder and Chief Executive Officer Elite Education Group “Topp” Jirayut Srupsrisopa Group Chief Executive Officer Bitkub Capital Group Holdings Co., Ltd Lady Mariéme Jamme Founder and Chief Executive Officer Amel Karboul iamtheCODE Founder and Chief Executive Officer Education Outcomes Fund Lasse Leponiemi Executive Director and Co-Founder Amit Patel HundrED Managing Director Owl Ventures Lydia Logan Vice-President, Global Education and Workforce Andreas Schleicher Development, Corporate Social Responsibility Director, Directorate of Education and Skills IBM Corporation Organisation for Economic Co-operation and Development (OECD) Maia Wagner Director, Global Impact Programs and Andria Zafirakou Community Management Teacher, Arts and Textile Dell Technologies Alperton Community School Neha Shah Asheesh Advani Co-Founder and President President and Chief Executive Officer GEP JA Worldwide Nina Huntemann Asif Saleh Chief Academic Officer Executive Director Chegg, Inc. BRAC Ninfa Salinas Sada Berith Bjørnholm Vice-President, Executive Committeer Senior Vice-President Grupo Salinas Novo Nordisk Foundation Njideka U. Harry Brian Johnsrud Founder and Member of the Executive Board Global Director, Education Learning and Advocacy Youth for Technology Foundation (YTF) Adobe Inc. Precious Moloi-Motsepe David Byer Co-Founder and Chief Executive Officer Strategic Advisor and Consultant Motsepe Foundation David Edwards Robert Jenkins General Secretary Director, Education and Adolescent Development, Education International United Nations Children’s Fund (UNICEF) Hadi Partovi Robert Palmer Founder and Chief Executive Officer Executive Director, Research and Programmes Code.org Queen Rania Foundation for Education and Development Heather Johnson Vice-President, Sustainability and Corporate Sherrie Westin Responsibility President Telefonaktiebolaget LM Ericsson Sesame Workshop Jamira Burley Valerie Singer Strategic Initiatives Lead, Worldwide Education General Manager, Global Education Apple Amazon Web Services Jean Daniel LaRock Wendy Kopp President and Chief Executive Officer Chief Executive Officer and Co-Founder Network for Teaching Entrepreneurship (NFTE) Teach For All Jeroo Billimoria Yejin Choi Founder Chief Executive Officer One Family Foundation DoBrain Shaping the Future of Learning: The Role of AI in Education 4.0 26  Endnotes 1. OECD (Organisation for Economic Co-operation and Development), PISA 2022 Results (Volume I): The State of Learning and Equity in Education, 2023. 2. OECD (Organisation for Economic Co-operation and Development), Future of Education and Skills 2030: OECD Learning Compass 2030, 2019. 3. World Economic Forum, Future of Jobs Report 2023, 2023. 4. United Nations, SDG4: Ensure inclusive and equitable quality education and promote lifelong learning opportunities for all, 2015. 5. UNESCO (United Nations Educational, Scientific, and Cultural Organization), The Teachers We Need for the Education We Want: The Global Imperative to Reverse the Teacher Shortage, 2023. 6. OECD (Organisation for Economic Co-operation and Development), Education at a Glance 2023, 2023, https://www. oecd-ilibrary.org/education/education-at-a-glance-2023_e13bef63-en. 7. A working paper released in November 2022 analysed four key aspects of the US K-12 teaching profession over the past 50 years: professional prestige, student interest, preparation for entry and job satisfaction. The findings reveal a bleak outlook, in particular a sharp decline in perceptions of teacher prestige, with drops ranging from 20% to 47% over the past decade. Interest in pursuing teaching careers among students has also plummeted drastically, declining by 50% since the 1990s and by 38% since 2010, reaching a 50-year low. Additionally, job satisfaction has significantly decreased; with the proportion of teachers feeling that the stress of their job is worthwhile, declining from 81% to 42% over the last 15 years. See: Kraft, Matthew A. and Melissa Arnold Lyon, The Rise and Fall of the Teaching Profession: Prestige, Interest, Preparation, and Satisfaction over the Last Half Century, EdWorking Paper No. 22-679, Providence: Annenberg Institute at Brown University, 2022, https:// edworkingpapers.com/sites/default/files/Kraft%20Lyon%202022%20State%20of%20the%20Teaching%20Profession_0.pdf. 8. The Open Innovation Team and the Government of the United Kingdom Department for Education, Generative AI in Education, 2023. 9. OECD (Organisation for Economic Co-operation and Development), Education at a Glance 2020, 2020. 10. OECD (Organisation for Economic Co-operation and Development), TALIS 2018 Results: Teachers and School Leaders as Valued Professionals, 2020. 11. Huang, Sonya, Pat Grady and GPT-3, Generative AI: A Creative New World, Sequoia Capital, 2022. 12. Deloitte, Talent and Workforce Effects in the Age of AI, 2020. 13. The Financial Times, Closing the AI skills gap, 2021. 14. World Economic Forum, “Education Meets AI”, World Economic Forum Annual Meeting, Davos-Klosters, 18 January 2024. 15. World Economic Forum, Jobs of Tomorrow: Large Language Models and Jobs, 2023. 16. Rand, David and Nathaniel Sirlin, “Digital Literacy Doesn’t Stop the Spread of Misinformation”, Scientific American, 15 July 2022, https://www.scientificamerican.com/article/digital-literacy-doesnt-stop-the-spread-of- misinformation/#:~:text=Someone%20lacking%20digital%20literacy%20skills,in%20the%20spread%20of%20 misinformation; UNESCO (United Nations Educational, Scientific, and Cultural Organization), Think Critically, Click Wisely: Media & Information Literacy Curriculum for Educators and Learners, Paris: UNESCO, 2021, https://unesdoc.unesco.org/ ark:/48223/pf0000377068/PDF/377068eng.pdf.multi. 17. Ewe, Koh, “The Ultimate Election Year: All the Elections Around the World in 2024”, Time, 28 December 2023. 18. ISTE (International Society for Technology in Education), Hands-on AI Projects for the Classroom, 2021. 19. Government of the United Kingdom Department for Education, Generative artificial intelligence (AI) in education, 2023, https:// www.gov.uk/government/publications/generative-artificial-intelligence-in-education/generative-artificial-intelligence-ai-in-education. 20. Bloom, Benjamin S., “The 2 Sigma Problem: The Search for Methods of Group Instruction as Effective as One-to-One Tutoring,” Educational Researcher, vol. 13, no. 6, 1984, pp. 4-16. 21. Cortes, Kalena, Karen Cortecamp, Susanna Loeb and Carly D. Robinson, Year Two Results Assessing the Effects of a Scalable Approach to High-Impact Tutoring for Young Readers, National Student Support Accelerator, 2023. 22. Fortune Business Insights, Market Research Report 2022, 2022. 23. Shemshack, Atika and Jonathan Michael Specter, “A Systemic Literature Review of Personalized Learning Terms”, Smart Learning Environments, vol. 7, no. 33, 2020. 24. Major, Louis, Gill A. Francis and Maria Tsapali, “The effectiveness of technology-supported personalized learning in low-and middle-income countries: A meta-analysis”, British Journal of Educational Technology, vol. 52, no. 5, 2020, pp. 1935-1964. 25. Hollands, Fiona and Venita Holmes, “How AI Tutoring Can Reshape Teachers’ Days, Education Week, 27 June 2023. 26. Gillespie, Nicole, Steven Lockey, Caitlin Curtis, Javad Pool and Ali Akbari, Trust in Artificial Intelligence: A Global Study, Brisbane: The University of Queensland, 2023. 27. ISACA, Generative AI: The Risks, Opportunities and Outlook, 2023. 28. World Economic Forum, EDISON Alliance Impact Report, 2024. 29. World Economic Forum, Global Risk Report 2024, 2024. Shaping the Future of Learning: The Role of AI in Education 4.0 27 The World Economic Forum, committed to improving the state of the world, is the International Organization for Public-Private Cooperation. The Forum engages the foremost political, business and other leaders of society to shape global, regional and industry agendas. World Economic Forum 91–93 route de la Capite CH-1223 Cologny/Geneva Switzerland Tel.: +41 (0) 22 869 1212 Fax: +41 (0) 22 786 2744 [email protected] www.weforum.org