How to navigate the future of AI in education and education in AI

A study conducted by EY teams surveyed 3,030 university students across eight regions and found that nearly half had used GenAI tools to support their studies, while an additional one-third expressed a willingness to use such resources. However, there were significant differences at the country level: 68% of students in Singapore and India were already using GenAI tools, compared to just 22% in Japan.

Similar student-led use cases were highlighted in the 2024 Digital Education Council Global AI Student Survey. The survey found that 68% of students used AI for research and 42% used it to check and improve their writing.

How can we enable AI to bring improved education outcomes for all?

AI has the potential to accelerate progress toward Sustainable Development Goal 4 (SDG 4), which aims to “ensure inclusive and equitable quality education and promote lifelong learning opportunities for all.” AI-powered tools can help tackle some of the critical challenges facing education systems worldwide. However, it is essential to develop and adopt AI tools in a way that genuinely enhances the learning experience without compromising the quality of education or academic integrity.

AI’s positive impact on education will not be equitable unless we address the digital divide. In many emerging economies, students and schools face barriers such as limited internet access, inadequate infrastructure, and a lack of digital skills. Bridging these gaps requires coordinated efforts, including investing in digital infrastructure, ensuring affordable internet and device access, and equipping educators with the skills to implement AI effectively. By addressing these disparities, we can promote AI-driven educational advancements that benefit students globally, creating a more equitable future where every learner has the opportunity to thrive.

Such initiatives need policies that promote equitable and inclusive access to digital technologies and AI. Additionally, these policies should encourage the ethical and inclusive use of AI for the public good.

The UN Educational, Scientific and Cultural Organization (UNESCO) suggests that “to fully unleash opportunities and mitigate potential risks, system-wide responses to the following key policy questions are needed:

• How can AI be leveraged to enhance education?
• How can we ensure the ethical, inclusive and equitable use of AI in education?
• How can education prepare humans to live and work with AI?”
  

According to an analysis by EY-Parthenon in February 2024, K-12 is the most popular market for AI tool development. The most common use cases in this sector are study tools and virtual tutors, which account for approximately 40% of tool development activity.

Supporting learning

AI tutors and teaching assistants provide personalized learning support by adapting to individual needs, learning styles, and preferences in real time. They offer immediate feedback and track progress, encouraging students to take ownership of their learning. By analyzing data from interactions, such as response times and accuracy, these tools suggest exercises, hints, or explanations tailored to each student's understanding. If a student masters a topic quickly, the AI presents more challenging material to maintain engagement.

Supporting teacher workflows

Many countries are experiencing shortage of teachers, with educators facing increasing workloads and burnout. AI tools can enhance efficiency in various job aspects, giving teachers more time to support student learning. For example, AI can assist with curriculum development, lesson planning, content creation, automated grading, feedback, and performance analytics, allowing teachers to focus on hands-on teaching.

A range of AI-driven learning management systems (LMS) is already available across education sectors, with established providers like Canvas, Blackboard Learn, and Moodle integrating AI features such as personalized learning paths and predictive analytics. This will lead to significant short-term adoption, particularly in K-12, as AI improves teacher workflows. By empowering teachers to become AI specialists, they can guide students in using AI responsibly, fostering an environment that emphasizes critical thinking and digital fluency.

Automating business processes and improving decision-making

Educational institutions are seeking to deploy AI to automate and streamline business processes in areas such as admissions, finance, human resources, marketing, communications and facilities management. This approach can enhance the experience of both employees and students while potentially leading to cost savings through increased productivity.

Examples include:

• Using AI models to identify students with a high likelihood of enrollment, improving offer-to-enrollment rates.
• Implementing an HR chatbot to provide a personalized hiring and onboarding experience or to respond to routine HR queries.
• Utilizing remote sensors and AI to manage heating and lighting, thereby reducing energy consumption.

Powering academic research

AI helps researchers by quickly scanning literature, identifying relevant papers, summarizing findings and highlighting research gaps. It analyzes large data sets to uncover patterns, creates synthetic data, and automates lab work, including experiments and data collection. Additionally, AI reduces the administrative burden of research by automating grant applications, compliance, reporting, and publishing.

Many universities globally are utilizing AI to enhance academic research across disciplines. For instance, Stanford University’s AI Laboratory (SAIL) conducts advanced research in AI, machine learning, and robotics while supporting AI applications in fields like health care, environmental science, and social sciences.

Enterprise and continuous learning

The innovation and adoption of AI in enterprise and continuous learning are rapidly accelerating. Unlike K-12 and higher education, where quality, accuracy, and security are paramount, enterprise learning has lower stakes and easier content creation. Enterprise users are more willing to take risks for potential rewards like cost savings and improved employee productivity.

Asynchronous digital learning solutions can transform enterprise learning and academic courses. These tools convert live course content into engaging, self-paced digital experiences, guided by the course owner. AI analyzes content elements to create interactive materials, such as reflection questions, games, and quizzes, eliminating the need for manual preparation. They can also recommend missing instructional items and produce accessible materials for learners with diverse needs or disabilities.

Policymakers and education leaders play a crucial role in mitigating risks and challenges associated with AI in education. They can achieve this by implementing robust policy frameworks. While adopting AI-driven digital tools has the potential to widen the educational divide, it also introduces various risks, including impacts on academic integrity, with concerns about plagiarism, cheating in assessments, and the loss of critical skills. Additionally, there are risks related to inaccuracy, discrimination, bias and data privacy.

Teachers will remain essential in education, even with AI-powered tutors. AI lacks the nuanced understanding and empathy of human educators, making it less effective in addressing complex learning needs. Additionally, AI tools have limitations in content recency and accuracy, often "hallucinating" or fabricating information. Students and teachers must learn to navigate these limitations and fact-check AI outputs. Consequently, AI tutors cannot yet be trusted to educate young children who may not recognize errors or hallucinations.

Transparency and explainability

AI models can function as "black boxes," with their inner workings often hidden or poorly understood. This lack of transparency complicates users' ability to understand the rationale behind decisions related to learning pathways, grading, or student performance, potentially introducing bias or errors. Without transparency and explainability, students, parents, and teachers cannot assess the fairness or accuracy of AI-based recommendations.

To promote widespread acceptance and effective use of AI in education, it's essential to establish protocols for model development that clarify how AI recommendations are generated and validated. Building transparency and explainability into AI models is vital for fostering confidence.

Privacy and data security

Privacy and data security are foundational for building confidence in AI-powered educational tools, promoting technology that enhances learning without compromising the safety and rights of students. In education, AI tools often rely on large amounts of student data to provide personalized learning experiences, track progress or adapt to individual needs However, collecting and storing sensitive information, such as academic performance and behavioral patterns, raises concerns about who has access to this data and how it can be used.

Impact on quality of learning and academic integrity

Educators are concerned that over-reliance on AI tools, particularly GenAI, may undermine students' skills in research, critical thinking, and effective communication. GenAI raises academic integrity issues, as it can produce essays, solve math problems, and complete tests that seem acceptable. Consequently, the focus has often been on prohibiting AI use or detecting AI-generated submissions. However, students need clear guidance and policies on the ethical and effective use of AI.

As schools and universities adopt AI-driven approaches, they should regularly monitor and evaluate outcomes while gathering stakeholder feedback, especially regarding equity. This enables necessary adjustments to enhance effectiveness and address any gaps.

Lack of digital and AI skills among staff and students

Most teachers have not yet integrated AI into their daily workflows, primarily due to a lack of knowledge on how to do so. They urgently need guidance, support, and resources to effectively use AI tools to enhance learning. A 2024 report by EdWeek Research Center found that the primary reason US teachers have not incorporated AI into their practice is a "lack of knowledge and support”.

Lack of digital infrastructure

To leverage AI in education, teachers and learners need access to digital devices, high-speed internet, and secure platforms. Without universal technology access, the digital divide and educational inequity will worsen, particularly in underserved communities lacking reliable internet and resources for computers. Promoting equitable AI-enhanced education requires investment in high-speed internet, connected devices, and, in some regions, stable electricity.

AI educational tools like adaptive learning platforms and intelligent tutoring systems require significant financial investment for setup, maintenance, updates, and data security. Implementing AI necessitates sophisticated data systems for managing student data, which must adhere to strict privacy standards, increasing costs for software, data protection, and staff training. Many schools, especially in underserved areas, struggle to fund traditional resources, let alone advanced technologies and their security needs.