The adoption of digital technology has resulted in many changes in education and learning. The set of basic skills that young people are expected to learn in school, at least in richer countries, has expanded to include a broad range of new ones to navigate the digital world. In many classrooms, paper has been replaced by screens and pens by keyboards. COVID-19 can be seen as a natural experiment where learning switched online for entire education systems virtually overnight. Higher education is the subsector with the highest rate of digital technology adoption, with online management platforms replacing campuses. The use of data analytics has grown in education management. Technology has made a wide range of informal learning opportunities accessible.

Yet the extent to which technology has transformed education needs to be debated. Change resulting from the use of digital technology is incremental, uneven and bigger in some contexts than others. The application of digital technology varies by community and socioeconomic level, by teacher willingness and preparedness, by education level, and by country income. Except in the most technologically advanced countries, computers and devices are not used in classrooms on a large scale. Technology use is not universal and will not become so any time soon. Moreover, evidence is mixed on its impact: Some types of technology seem to be effective in improving some kinds of learning. The short- and long-term costs of using digital technology appear to be significantly underestimated. The most disadvantaged are typically denied the opportunity to benefit from this technology.

Too much attention on technology in education usually comes at a high cost. Resources spent on technology, rather than on classrooms, teachers and textbooks for all children in low- and lower-middle-income countries lacking access to these resources are likely to lead to the world being further away from achieving the global education goal, SDG 4. Some of the world’s richest countries ensured universal secondary schooling and minimum learning competencies before the advent of digital technology. Children can learn without it.

However, their education is unlikely to be as relevant without digital technology. The Universal Declaration of Human Rights defines the purpose of education as promoting the ‘full development of the human personality’, strengthening ‘respect for … fundamental freedoms’ and promoting ‘understanding, tolerance and friendship’. This notion needs to move with the times. An expanded definition of the right to education could include effective support by technology for all learners to fulfil their potential, regardless of context or circumstance.

Clear objectives and principles are needed to ensure that technology use is of benefit and avoids harm. The negative and harmful aspects in the use of digital technology in education and society include risk of distraction and lack of human contact. Unregulated technology even poses threats to democracy and human rights, for instance through invasion of privacy and stoking of hatred. Education systems need to be better prepared to teach about and through digital technology, a tool that must serve the best interests of all learners, teachers and administrators. Impartial evidence showing that technology is being used in some places to improve education, and good examples of such use, need to be shared more widely so that the optimal mode of delivery can be assured for each context.

• Equity and inclusion: Is fulfilment of the right to choose the education one wants and to realize one’s full potential through education compatible with the goal of equality? If not, how can education become the great equalizer?
• Quality: Do education’s content and delivery support societies in achieving sustainable development objectives? If not, how can education help learners to not only acquire knowledge but also be agents of change?
• Efficiency: Does the current institutional arrangement of teaching learners in classrooms support the achievement of equity and quality? If not, how can education balance individualized instruction and socialization needs?

How best can digital technology be included in a strategy to tackle these challenges, and under what conditions? Digital technology packages and transmits information on an unprecedented scale at high speed and low cost. Information storage has revolutionized the volume of accessible knowledge. Information processing enables learners to receive immediate feedback and, through interaction with machines, adapt their learning pace and trajectory: Learners can organize the sequence of what they learn to suit their background and characteristics. Information sharing lowers the cost of interaction and communication. But while such technology has tremendous potential, many tools have not been designed for application to education. Not enough attention has been given to how they are applied in education and even less to how they should be applied in different education contexts.

On the question of equity and inclusion, ICT – and digital technology in particular – helps lower the education access cost for some disadvantaged groups: Those who live in remote areas are displaced, face learning difficulties, lack time or have missed out on past education opportunities. But while access to digital technology has expanded rapidly, there are deep divides in access. Disadvantaged groups own fewer devices, are less connected to the internet (Figure 1) and have fewer resources at home. The cost of much technology is falling rapidly but is still too high for some. Households that are better off can buy technology earlier, giving them more advantages and compounding disparity. Inequality in access to technology exacerbates existing inequality in access to education, a weakness exposed during the COVID-19 school closures.

Figure 1: Internet connectivity is highly unequal

Percentage of 3- to 17-year-olds with internet connection at home, by wealth quintile, selected countries, 2017–19
Source: UNICEF database.

Education quality is a multifaceted concept. It encompasses adequate inputs (e.g. availability of technology infrastructure), prepared teachers (e.g. teacher standards for technology use in classrooms), relevant content (e.g. integration of digital literacy in the curriculum) and individual learning outcomes (e.g. minimum levels of proficiency in reading and mathematics). But education quality should also encompass social outcomes. It is not enough for students to be vessels receiving knowledge; they need to be able to use it to help achieve sustainable development in social, economic and environmental terms.

There are a variety of views on the extent to which digital technologies can enhance education quality. Some argue that, in principle, digital technology creates engaging learning environments, enlivens student experiences, simulates situations, facilitates collaboration and expands connections. But others say digital technology tends to support an individualized approach to education, reducing learners’ opportunities to socialize and learn by observing each other in real-life settings. Moreover, just as new technology overcomes some constraints, it brings its own problems. Increased screen time has been associated with adverse impact on physical and mental health. Insufficient regulation has led to unauthorized use of personal data for commercial purposes. Digital technology has also helped spread misinformation and hate speech, including through education.

Improvements to efficiency may be the most promising way for digital technology to make a difference in education. Technology is touted as being able to reduce the time students and teachers spend on menial tasks, time that can be used in other, educationally more meaningful activities. However, there are conflicting views on what is meaningful. The way that education technology is used is more complex than just a substitution of resources. Technology may be one-to-many, one-to-one or peer-to-peer technology. It may require students to learn alone or with others, online or offline, independently or networked. It delivers content, creates learner communities and connects teachers with students. It provides access to information. It may be used for formal or informal learning and can assess what has been learned. It is used as a tool for productivity, creativity, communication, collaboration, design and data management. It may be professionally produced or have user-generated content. It may be specific to schools and place-based or transcend time and place. As in any complex system, each technology tool involves distinct infrastructure, design, content and pedagogy, and each may promote different types of learning.

Technology is evolving too fast to permit evaluation that could inform decisions on legislation, policy and regulation. Research on technology in education is as complex as technology itself. Studies evaluate experiences of learners of various ages using various methodologies applied in contexts as different as self-study, classrooms and schools of diverse sizes and features, non-school settings, and at system level. Findings that apply in some contexts are not always replicable elsewhere. Some conclusions can be drawn from long-term studies as technologies mature but there is an endless stream of new products. Meanwhile, not all impact can be easily measured, given technology’s ubiquity, complexity, utility and heterogeneity. In brief, while there is much general research on education technology, the amount of research for specific applications and contexts is insufficient, making it difficult to prove that a particular technology enhances a particular kind of learning.

Why is there often the perception nevertheless that technology can address major education challenges? To understand the discourse around education technology, it is necessary to look behind the language being used to promote it, and the interests it serves. Who frames the problems technology should address? What are the consequences of such framing for education? Who promotes education technology as a precondition for education transformation? How credible are such claims? What criteria and standards need to be set to evaluate digital technology’s current and potential future contribution to education so as to separate hype from substance? Can evaluation go beyond short-term assessments of impact on learning and capture potential far-reaching consequences of the generalized use of digital technology in education?

A wide range of technology brings education to hard-to-reach learners. Technology has historically opened up education to learners facing significant obstacles in access to schools or well-trained teachers. Interactive radio instruction is used in nearly 40 countries. In Nigeria, radio instruction combined with print and audiovisual materials has been used since the 1990s, reaching nearly 80% of nomads and increasing their literacy, numeracy and life skills. Television has helped educate marginalized groups, notably in Latin America and the Caribbean. The Telesecundaria programme in Mexico, combining televised lessons with in-class support and extensive teacher training, increased secondary school enrolment by 21%. Mobile learning devices, often the only type of device accessible to disadvantaged learners, have been used in hard-to-reach areas and emergencies to share educational materials; complement in-person or remote channels; and foster interactions between students, teachers and parents, notably during COVID-19. Adults have been the main target of online distance learning, with open universities having increased participation for both working and disadvantaged adults.

Inclusive technology supports accessibility and personalization for learners with disabilities. Assistive technology removes learning and communication barriers, with numerous studies reporting a significant positive impact on academic engagement, social participation and the well-being of learners with disabilities. However, such devices remain inaccessible and unaffordable in many countries, and teachers often lack specialized training to use them effectively in learning environments. While people with disabilities used to rely exclusively on specialized devices to gain access to education, technology platforms and devices are increasingly incorporating accessibility features, which support inclusive, personalized learning for all students.

Technology supports learning continuity in emergencies. Mapping of 101 distance education projects in crisis contexts in 2020 showed that 70% used radio, television and basic mobile phones. During the Boko Haram crisis in Nigeria, the Technology Enhanced Learning for All programme used mobile phones and radios to support the learning continuity of 22,000 disadvantaged children, with recorded improvement in literacy and numeracy skills. However, there are significant gaps in terms of rigorous evaluation of education technology in emergencies, despite some limited recorded impact. Meanwhile, most projects are led by non-state actors as short-term crisis responses, raising sustainability concerns; education ministries implemented only 12% of the 101 projects.

Technology supported learning during COVID-19, but millions were left out. During school closures, 95% of education ministries carried out some form of distance learning, potentially reaching over 1 billion students globally. Many of the resources used during the pandemic were first developed in response to previous emergencies or rural education, with some countries building on decades of experience with remote learning. Sierra Leone revived the Radio Teaching Programme, developed during the Ebola crisis, one week after schools closed. Mexico expanded content from its Telesecundaria programme to all levels of education. However, at least half a billion, or 31% of students worldwide – mostly the poorest (72%) and those in rural areas (70%) – could not be reached by remote learning. Although 91% of countries used online learning platforms to deliver distance learning during school closures, the platforms only reached a quarter of students globally. For the rest, low-tech interventions such as radio and television were largely used, in combination with paper-based materials and mobile phones for increased interactivity.

Some countries are expanding existing platforms to reach marginalized groups. Less than half of all countries developed long-term strategies for increasing their resilience and the sustainability of interventions as part of their COVID-19 response plans. Many have abandoned distance learning platforms developed during COVID-19, while others are repurposing them to reach marginalized learners. The digital platform set up in Ukraine during the pandemic was expanded once the war broke out in 2022, allowing 85% of schools to complete the academic year.

Technology facilitates content creation and adaptation. Open educational resources (OERs) encourage the reuse and repurposing of materials to cut development time, avoid duplication of work and make materials more context-specific or relevant to learners. They also significantly reduce the cost of access to content. In the US state of North Dakota, an initial investment of USD 110,000 to shift to OERs led to savings of over USD 1 million in student costs. Social media increases access to user-generated content. YouTube, a major player in both formal and informal learning, is used by about 80% of the world’s top 113 universities. Moreover, collaborative digital tools can improve the diversity and quality of content creation. In South Africa, the Siyavule initiative supported tutor collaboration on the creation of primary and secondary education textbooks.

Digitization of educational content simplifies access and distribution. Many countries, including Bhutan and Rwanda, have created static digital versions of traditional textbooks to increase availability. Others, including India and Sweden, have produced digital textbooks that encourage interactivity and multimodal learning. Digital libraries and educational content repositories such as the National Academic Digital Library of Ethiopia, National Digital Library of India and Teachers Portal in Bangladesh help teachers and learners find relevant materials. Learning management platforms, which have become a key part of the contemporary learning environment, help organize content by integrating digital resources into course structures.

Open access resources help overcome barriers. Open universities and MOOCs can eliminate time, location and cost barriers to access. In Indonesia, where low participation in tertiary education is largely attributed to geographical challenges, MOOCs play an important role in expanding access to post-secondary learning. During COVID-19, MOOC enrolment surged, with the top three providers adding as many users in April 2020 as in all of 2019. Technology can also remove language barriers. Translation tools help connect teachers and learners from various countries and increase the accessibility of courses by non-native students.

Ensuring and assessing the quality of digital content is difficult. The sheer quantity of content and its decentralized production pose logistical challenges for evaluation. Several strategies have been implemented to address this. China established specific quality criteria for MOOCs to be nationally recognized. The European Union developed its OpenupED quality label. India strengthened the link between non-formal and formal education. Micro-credentials are increasingly used to ensure that institution and learner both meet minimum standards. Some platforms aim to improve quality by recentralizing content production. YouTube, for example, has been funnelling financing and resources to a few trusted providers and partnering with well-established education institutions.

Technology may reinforce existing inequality in both access to and production of content. Privileged groups still produce most content. A study of higher-education repositories with OER collections found that nearly 90% were created in Europe or North America; 92% of the material in the OER Commons global library is in English. This influences who has access to digital content. MOOCs, for example, mainly benefit educated learners – studies have shown around 80% of participants on major platforms already have a tertiary degree – and those from richer countries. The disparity is due to divides in digital skills, internet access, language and course design. Regional MOOCs cater to local needs and languages but can also worsen inequality.

The prevalence of ICT use in classrooms is not high, even in the world’s richest countries. The 2018 PISA found that only about 10% of 15-year-old students in over 50 participating education systems used digital devices for more than an hour a week in mathematics and science lessons, on average (Figure 2). The 2018 International Computer and Information Literacy Study (ICILS) showed that in the 12 participating education systems, simulation and modelling software in classrooms was available to just over one third of students, with country levels ranging from 8% in Italy to 91% in Finland.

Figure 2: Even in upper-middle- and high-income countries, technology use in mathematics and science classrooms
is limited

Percentage of 15-year-old students who used digital devices for at least one hour per week in mathematics or science classroom lessons, selected upper-middle- and high-income countries, 2018
Source: 2018 PISA database.

Recorded lessons can address teacher quality gaps and improve teacher time allocation. In China, lesson recordings from high-quality urban teachers were delivered to 100 million rural students. An impact evaluation showed improvements in Chinese skills by 32% and a 38% long-term reduction in the rural–urban earning gap. However, just delivering materials without contextualizing and providing support is insufficient. In Peru, the One Laptop Per Child programme distributed over 1 million laptops loaded with content, but no positive impact on learning resulted, partly due to the focus on provision of devices instead of the quality of pedagogical integration.

Enhancing technology-aided instruction with personalization can improve some types of learning. Personalized adaptive software generates analytics that can help teachers track student progress, identify error patterns, provide differentiated feedback and reduce workload on routine tasks. Evaluations of the use of a personalized adaptive software in India documented learning gains in after-school settings and for low-performing students. However, not all widely used software interventions have strong evidence of positive effects compared to teacher-led instruction. A meta-analysis of studies on an AI learning and assessment system that has been used by over 25 million students in the United States found it was no better than traditional classroom teaching in improving outcomes.

Varied interaction and visual representation can enhance student engagement. A meta-analysis of 43 studies published from 2008 to 2019 found that digital games improved cognitive and behavioural outcomes in mathematics. Interactive whiteboards can support teaching and learning if well integrated in pedagogy; but in the United Kingdom, despite large-scale adoption, they were mostly used to replace blackboards. Augmented, mixed or virtual reality used as an experiential learning tool for repeated practice in life-like conditions in technical, vocational and scientific subjects is not always as effective as real-life training but may be superior to other digital methods, such as video demonstrations.

Technology offers teachers low-cost and convenient ways to communicate with parents. The Colombian Institute of Family Welfare’s distance education initiative, which targeted 1.7 million disadvantaged children, relied on social media platforms to relay guidance to caregivers on pedagogical activities at home. However, uptake and effectiveness of behavioural interventions targeting caregivers are limited by parental education levels, as well as lack of time and material resources.

Student use of technology in classrooms and at home can be distracting, disrupting learning. A meta-analysis of research on student mobile phone use and its impact on education outcomes, covering students from pre-primary to higher education in 14 countries, found a small negative effect, and a larger one at the university level. Studies using PISA data indicate a negative association between ICT use and student performance beyond a threshold of moderate use. Teachers perceive tablet and phone use as hampering classroom management. More than one in three teachers in seven countries participating in the 2018 ICILS agreed that ICT use in classrooms distracted students. Online learning relies on student ability to self-regulate and may put low-performing and younger learners at increased risk of disengagement.

The definition of digital skills has been evolving along with digital technology. An analysis for this report shows that 54% of countries have identified digital skills standards for learners. The Digital Competence Framework for Citizens (DigComp), developed on behalf of the European Commission, has five competence areas: information and data literacy, communication and collaboration, digital content creation, safety, and problem-solving. Some countries have adopted digital skills frameworks developed by non-state, mostly commercial, actors. The International Computer Driving Licence (ICDL) has been promoted as a ‘digital skills standard’ but is associated mainly with Microsoft applications. Kenya and Thailand have endorsed the ICDL as the digital literacy standard for use in schools.

Digital skills are unequally distributed. In the 27 European Union (EU) countries, 54% of adults had at least basic digital skills in 2021. In Brazil, 31% of adults had at least basic skills, but the level was twice as high in urban as in rural areas, three times as high among those in the labour force as among those outside it, and nine times as high in the top socioeconomic group as in the two bottom groups. The overall gender gap in digital skills is small, but wider in specific skills. In 50 countries, 6.5% of males and 3.2% of females could write a computer program. In Belgium, Hungary and Switzerland, no more than 2 women for every 10 men could program; in Albania, Malaysia and Palestine, 9 women for every 10 men could do so. According to the 2018 PISA, 5% of 15-year-olds with the strongest reading skills but 24% of those with the weakest ones were at risk of being misled by a typical phishing email.

Formal skills training may not be the main way of acquiring digital skills. About one quarter of adults in EU countries, ranging from 16% in Italy to 40% in Sweden, had acquired skills through a ‘formalised educational institution’. Informal learning, such as self-study and informal assistance from colleagues, relatives and friends, was used by twice as many. Still, formal education is important: In 2018, those with tertiary education in Europe were twice as likely (18%) as those with upper secondary education (9%) to engage in free online training or self-study to improve their computer, software or application use. Solid mastery of literacy and numeracy skills is positively associated with mastery of at least some digital skills.

A curriculum content mapping of 16 education systems showed that Greece and Portugal dedicated less than 10% of the curriculum to data and media literacy while Estonia and the Republic of Korea embedded both in half their curricula. In some countries, media literacy in curricula is explicitly connected to critical thinking in subject disciplines, as under Georgia’s New School Model. Asia is characterized by a protectionist approach to media literacy that prioritizes information control over education. But in the Philippines, the Association for Media and Information Literacy successfully advocated for incorporation of media and information literacy in the curriculum, and it is now a core subject in grades 11 and 12.

Digital skills in communication and collaboration matter in hybrid learning arrangements. Argentina promoted teamwork skills as part of a platform for programming and robotics competitions in primary and secondary education. Mexico offers teachers and students digital education resources and tools for remote collaboration, peer learning and knowledge sharing. Ethical digital behaviour includes rules, conventions and standards to be learned, understood and practised by digital users when using digital spaces. Digital communication’s anonymity, invisibility, asynchronicity and minimization of authority can make it difficult for individuals to understand its complexities.

Competences in digital content creation include selecting appropriate delivery formats and creating copy, audio, video and visual assets; integrating digital content; and respecting copyright and licences. The ubiquitous use of social media has turned content creation into a skill with direct application in electronic commerce. In Indonesia, the Siberkreasi platform counts collaborative engagement among its core activities. The Kenya Copyright Board collaborates closely with universities to provide copyright education and conducts frequent training sessions for students in the visual arts and ICT.

Education systems need to strengthen preventive measures and respond to many safety challenges, from passwords to permissions, helping learners understand the implications of their online presence and digital footprint. In Brazil, 29% of schools have conducted debates or lectures on privacy and data protection. In New Zealand, the Te Mana Tūhono (Power of Connectivity) programme delivers digital protection and security services to almost 2,500 state and state-integrated schools. A systematic review of interventions in Australia, Italy, Spain and the United States estimated that the average programme had a 76% chance of reducing cyberbullying perpetration. In Wales, United Kingdom, the government has advised schools how to prepare for and respond to harmful viral online content and hoaxes.

The definition of problem-solving skills varies widely among education systems. Many countries perceive them in terms of coding and programming and as part of a computer science curriculum that includes computational thinking, algorithm use and automation. A global review estimated that 43% of students in high-income countries, 62% in upper-middle-income, 5% in lower-middle-income but no students in low-income countries take computer science as compulsory in primary and/or secondary education. Only 20% of education systems require schools to offer computer science as an elective or core course. Non-state actors often support coding and programming skills. In Chile, Code.org has partnered with the government to provide educational resources in computer science.

Education management information systems focus on efficiency and effectiveness. Education reforms have been characterized by increased school autonomy, target setting and results-based performance, all of which require more data. By one measure, since the 1990s, the number of policies making reference to data, statistics and information has increased by 13 times in high-income, 9 times in upper-middle-income, and 5 times in low- and lower-middle-income countries. But only 54% of countries globally – and as low as 22% in sub-Saharan Africa – have unique student identification mechanisms.

Geospatial data can support education management. Geographical information systems help address equity and efficiency in infrastructure and resource distribution in education systems. School mapping has been used to foster diversity and reduce inequality of opportunity. Ireland links three databases to decide in which of its 314 planning areas to build new schools. Geospatial data can identify areas where children live too far from the nearest school. For instance, it has been estimated that 5% of the population in Guatemala and 41% in the United Republic of Tanzania live more than 3 kilometres away from the nearest primary school.

Education management information systems struggle with data integration. In 2017, Malaysia introduced the Education Data Repository as part of its 2019–23 ICT Transformation Plan to progressively integrate its 350 education data systems and applications scattered across institutions. By 2019, it had integrated 12 of its main data systems, aiming for full integration through a single data platform by the end of 2023. In New Zealand, schools had been procuring student management systems independently and lack of interoperability between them was preventing authorities from tracking student progress. In 2019, the government began setting up the National Learner Repository and Data Exchange to be hosted in cloud data centres, but deployment was paused in 2021 due to cybersecurity concerns. European countries have been addressing interoperability concerns collectively to facilitate data sharing between countries and across multiple applications used in higher-education management through the EMREX project.

Computer-based assessments and computer adaptive testing have been replacing many paper-based assessments. They reduce test administration costs, improve measurement quality and provide rapid scoring. As more examinations shift online, the need for online cheating detection and proctoring tools has also increased. While these can reduce cheating, their effectiveness should be weighed against fairness and psychological effects. Evidence on the quality and usefulness of technology-based assessments has started to emerge, but much less is known about cost efficiency. Among 34 papers on technology-based assessments reviewed for this report, transparent data on cost were lacking.

Learning analytics can increase formative feedback and enable early detection systems. In China, learning analytics has been used to identify learners’ difficulties, predict learning trajectories and manage teacher resources. In the United States, Course Signals is a system used to flag the likelihood of a student not passing a course; educators can then target them for additional support. However, learning analytics requires all actors to have sufficient data literacy. Successful education systems typically have absorptive capacity, including strong school leaders and confident teachers willing to innovate. Yet often seemingly trivial issues, such as maintenance and repair, are ignored or underestimated.

Access to electricity and devices is highly unequal between and within countries. In 2021, almost 9% of the global population – and more than 70% of people in rural sub-Saharan Africa – lacked access to electricity. Globally, one in four primary schools do not have electricity. A 2018 study in Cambodia, Ethiopia, Kenya, Myanmar, Nepal and Niger found that 31% of public schools were on grid and 9% were off grid, with only 16% enjoying uninterrupted power supply. Globally, 46% of households had a computer at home in 2020; the share of schools with computers for pedagogical purposes was 47% in primary, 62% in lower secondary and 76% in upper secondary education. There were at most 10 computers per 100 students in Brazil and Morocco but 160 computers per 100 students in Luxembourg, according to the 2018 PISA.

Internet access, a vital enabler of economic, social and cultural rights, is also unequal. In 2022, two in three people globally used the internet. In late 2021, 55% of the world’s population had mobile broadband access. In low- and middle-income countries, 16% less women than men used mobile internet in 2021. An estimated 3.2 billion people do not use mobile internet services despite being covered by a mobile broadband network. Globally, 40% of primary, 50% of lower secondary and 65% of upper secondary schools are connected to the internet. In India, 53% of private unaided and 44% of private aided schools are connected, compared with only 14% of government schools.

Various policies are used to improve access to devices. Some one in five countries have policies granting subsidies or deductions to buy devices. One-to-one technology programmes were established in 30% of countries at one time; currently only 15% of countries pursue such programmes. A number of upper-middle- and high-income countries are shifting from providing devices to allowing students to use their own devices in school. Jamaica adopted a Bring Your Own Device policy framework in 2020 to aim for sustainability.

Some countries champion free and open source software. Education institutions with complex ICT infrastructure, such as universities, can benefit from open source software to add new solutions or functionalities. By contrast, proprietary software does not permit sharing and has vendor locks that hinder interoperability, exchange and updates. In India, the National e-Governance Plan makes it mandatory for all software applications and services used in government to be built on open source software to achieve efficiency, transparency, reliability and affordability.

Countries are committed to universal internet provision at home and in school. About 85% of countries have policies to improve school or learner connectivity and 38% have laws on universal internet provision. A review of 72 low- and middle-income countries found that 29 had used universal service funds to reduce costs for underserved groups. In Kyrgyzstan, renegotiated contracts helped cut prices by nearly half and almost doubled internet speed. In Costa Rica, the Hogares Conectados (Connected Households) programme, which provided an internet cost subsidy to the poorest 60% of households with school-age children, helped reduce the share of unconnected households from 41% in 2016 to 13% in 2019. Zero-rating, or providing free internet access for education or other purposes, has been used, especially during COVID-19, but is not without problems, as it violates the net neutrality principle.

Education technology is often underutilized. In the United States, an average of 67% of education software licences were unused and 98% were not used intensively. According to the EdTech Genome Project, 85% of some 7,000 pedagogical tools, which cost USD 13 billion, were ‘either a poor fit or implemented incorrectly’. Less than one in five of the top 100 education technology tools used in classrooms met the requirements of the US Every Student Succeeds Act. Research had been published for 39% of these tools but the research was aligned with the act in only 26% of cases.

Evidence needs to drive education technology decisions. A review in the United Kingdom found that only 7% of education technology companies had conducted randomized controlled trials, 12% had used third-party certification and 18% had engaged in academic studies. An online survey of teachers and administrators in 17 US states showed that only 11% requested peer-reviewed evidence prior to adopting education technology. Recommendations influence purchase decisions, yet ratings can be manipulated through fake reviews disseminated on social media. Few governments try to fill the evidence gap, so demand has grown for independent reviews. Edtech Tulna, a partnership between a private think tank and a public university in India, offers quality standards, an evaluation toolkit and publicly available expert reviews.

Education technology procurement decisions need to take economic, social and environmental sustainability into account. With respect to economic considerations, it is estimated that initial investment in education technology accounts for just 25% or less of the eventual total cost. Regarding social concerns, procurement processes need to address equity, accessibility, local ownership and appropriation. In France, the Territoires Numériques Educatifs (Digital Educational Territories) initiative was criticized because not all subsidized equipment met local needs, and local governments were left out of the decisions on which equipment to purchase. Both issues have since been addressed. Concerning environmental considerations, it has been estimated that extending the lifespan of all laptops in the European Union by a year would save the equivalent of taking almost 1 million cars off the road in terms of CO2 emissions.

Regulation needs to address risks in education technology procurement. Public procurement is vulnerable to collusion and corruption. In 2019, Brazil’s Comptroller General of the Union found irregularities in the electronic bidding process for the purchase of 1.3 million computers, laptops and notebooks for state and municipal public schools. Decentralizing public procurement to local governments is one way to balance some of the risks. Indonesia has used its SIPLah e-commerce platform to support school-level procurement processes. However, decentralization is vulnerable to weak organizational capacity. A survey of administrators in 54 US school districts found that they had rarely carried out needs assessments.

Governance of the education technology system is fragmented. A department or an agency responsible for education technology has been identified in 82% of countries. Placing education ministries in charge of education technology strategies and plans could help ensure that decisions are primarily based on pedagogical principles. However, this is the case in just 58% of countries. In Kenya, the 2019 National Information, Communications and Technology Policy led the Ministry of Information, Communications and Technology to integrate ICT at all levels of education.

Participation is often limited in the development of education technology strategies and plans. Nepal established a Steering and a Coordination Committee under the 2013–17 ICT in Education Master Plan for intersectoral and inter-agency coordination and cooperation in its implementation. Including administrators, teachers and students can help bridge the knowledge gap with decision makers to ensure that education technology choices are appropriate. In 2022, only 41% of US education sector leaders agreed that they were regularly included in planning and strategic conversations about technology.

The private sector’s commercial interests can clash with government equity, quality and efficiency goals. In India, the government alerted families about the hidden costs of free online content. Other risks relate to data use and protection, privacy, interoperability and lock-in effects, whereby students and teachers are compelled to use specific software or platforms. Google, Apple and Microsoft produce education platforms tied to particular hardware and operating systems.

Privacy risks to children make their learning environment unsafe. One analysis found that 89% of 163 education technology products recommended for children’s learning during the COVID-19 pandemic could or did watch children outside school hours or education settings. In addition, 39 of 42 governments providing online education during the pandemic fostered uses that ‘risked or infringed’ upon children’s rights. Data used for predictive algorithms can bias predictions and decisions and lead to discrimination, privacy violations and exclusion of disadvantaged groups. The Cyberspace Administration of China and the Ministry of Education introduced regulations in 2019 requiring parental consent before devices powered by AI, such as cameras and headbands, could be used with students in schools and required data to be encrypted.

Children’s exposure to screen time has increased. A survey of screen time of parents of 3- to 8-year-olds in Australia, China, Italy, Sweden and the United States found that their children’s screen exposure increased by 50 minutes during the pandemic for both education and leisure. Extended screen time can negatively affect self-control and emotional stability, increasing anxiety and depression. Few countries have strict regulations on screen time. In China, the Ministry of Education limited the use of digital devices as teaching tools to 30% of overall teaching time. Less than one in four countries are banning the use of smartphones in schools. Italy and the United States have banned the use of specific tools or social media from schools. Cyberbullying and online abuse are rarely defined as offences but can fall under existing laws, such as stalking laws as in Australia and harassment laws in Indonesia.

Monitoring of data protection law implementation is needed. Only 16% of countries explicitly guarantee data privacy in education by law and 29% have a relevant policy, mainly in Europe and Northern America. The number of cyberattacks in education is rising. Such attacks increase exposure to theft of identity and other personal data, but capacity and funds to address the issue are often insufficient. Globally, 5% of all ransomware attacks targeted the education sector in 2022, accounting for more than 30% of cybersecurity breaches. Regulations on sharing children’s personal information are rare but are starting to emerge under the EU’s General Data Protection Regulation. China and Japan have binding instruments on protecting children’s data and information.

Education systems support teachers in developing technology-related professional competencies. About half of education systems worldwide have ICT standards for teachers in a competency framework, teacher training framework, development plan or strategy. Education systems set up annual digital education days for teachers, promote OER, support the exchange of experiences and resources between teachers, and offer training. One quarter of education systems have legislation to ensure teachers are trained in technology, either through initial or in-service training. Some 84% of education systems have strategies for in-service teacher professional development, compared with 72% for pre-service teacher education in technology. Teachers can identify their development needs using digital self-assessment tools such as that provided by the Centre for Innovation in Brazilian Education.

Technology is changing teacher training. Technology is used to create flexible learning environments, engage teachers in collaborative learning, support coaching and mentoring, increase reflective practice, and improve subject or pedagogical knowledge. Distance education programmes have promoted teacher learning in South Africa and even equalled the impact of in-person training in Ghana. Virtual communities have emerged, primarily through social networks, for communication and resource sharing. About 80% of teachers surveyed in the Caribbean belonged to professional WhatsApp groups and 44% used instant messaging to collaborate at least once a week. In Senegal, the Reading for All programme used in-person and online coaching. Teachers considered face-to-face coaching more useful, but online coaching cost 83% less and still achieved a significant, albeit small, improvement in how teachers guided students’ reading practice. In Flanders, Belgium, KlasCement, a teacher community network created by a non-profit and now run by the Ministry of Education, expanded access to digital education and provided a platform for discussions on distance education during the pandemic.

Many actors support teacher professional development in ICT. Universities, teacher training institutions and research institutes provide specialized training, research opportunities and partnerships with schools for professional development in ICT. In Rwanda, universities collaborated with teachers and the government to develop the ICT Essentials for Teachers course. Teacher unions also advocate for policies that support teachers. The Confederation of Education Workers of the Argentine Republic established the right of teachers to disconnect. Civil society organizations, including the Carey Institute for Global Good, offer support through initiatives such as providing OER and online courses for refugee teachers in Chad, Kenya, Lebanon and Niger.