How AI Technology Can Advance Higher Education

As China works to build a strong education system, how to advance the “AI Plus” initiative has become a major concern for educators. Higher education urgently needs profound changes in its educational philosophy, institutional development, teaching models, learning paradigms and assessment methods to shape a new ecosystem of AI-enhanced higher education.

At what stage is the application of AI in higher education, and where will it develop further? What value does AI bring to educating students and transforming education? Professor Huang Ronghuai, UNESCO Chair on AI in Education and Director of the Ministry of Education’s Education Informatization Strategy Research Base (Beijing), answers these questions about how AI technology can advance higher education.

What changes is higher education undergoing with the support of AI technology?

In response to the new wave of AI development, higher education is undergoing systemic changes in learning environments, teaching, assessment and research, actively adapting to the changing development of an intelligent society. These changes include the following.

Connected and integrated intelligent learning environments are taking shape.

AI, the Internet of Things and other key technologies connect people, machines and environments. New spaces combining comprehensive sensing, integrated physical and virtual connectivity, and intelligent interaction are becoming a direction for the next generation of learning environments. For example, the Ministry of Education is exploring pilot future learning centres that draw on university libraries, integrate institutions’ learning resources and use AI to deliver learning services on demand, efficiently and precisely, thereby transforming learning spaces.

Human–AI collaborative forms of teaching are emerging.

Higher education teaching in the intelligent era is entering a setting of multiple forms of collaboration, where human teachers and intelligent machines coexist and teach together. Using machines’ strengths to complement teachers’ limitations can improve efficiency and quality while meeting diverse teaching needs. For example, the interview describes the Massachusetts Institute of Technology’s exploration of AI to help teachers improve their methods and efficiency, and its use of generative-AI-driven personalised learning systems in engineering and computer science courses. These systems adjust content to students’ performance and learning styles to help them understand complex concepts.

Dynamic and multidimensional assessment is developing.

Cross-context educational assessment platforms supported by AI and big data are becoming an important part of universities’ digital campuses. Assessment is evolving from a single approach centred on scores and final outcomes towards technology-supported, data-driven, multidimensional assessment of the learning process. For example, Beijing Normal University has explored intelligent “AI Plus” classroom assessment. Combining computer vision, natural language processing, ensemble learning and statistical modelling, its system assesses classroom teaching processes to overcome the subjectivity and delay of traditional evaluation, substantially reduce labour costs and monitor teaching processes more precisely.

Interdisciplinary and innovative approaches to research are taking shape.

AI’s capabilities offer new ways to address research problems. They are accelerating research efficiency, advancing new research paradigms and changing the form of research in higher education.

At what stage is AI application in higher education, and where should further development focus?

With its responsive adaptability and deep research foundations, higher education is on the eve of profound educational change. Yet even as technology enables innovation and transformation, the education system continues to face tensions on several fronts.

Five tensions to address

1. Capital-driven development versus education’s internal logic. Capital accelerates the concentration of resources and the spread of technologies such as AI in education, but its pursuit of rapid returns can distort education’s public-interest character and long-term development goals.
2. Widespread technology use versus flexibility in assessment. Students need to embrace and adapt to rapid technological development. However, the convenience of technology can also encourage excessive dependence or misuse, such as using AI to complete assignments or cheat in examinations, challenging assessment systems.
3. Standardised education versus students’ developmental needs. The mainstream model for developing talent still follows industrial-era standardised education. Teaching consequently revolves around mastering basic knowledge and skills, making it difficult to meet students’ needs for well-rounded development.
4. Established teaching habits versus innovation in teaching methods. When faced with large university classes, many teachers struggle to accommodate differences among students and are likely to continue using familiar traditional methods.
5. Programme provision versus societal needs. As AI rapidly evolves and society’s needs keep changing, universities often lag in establishing relevant disciplines and updating course content. This creates a mismatch between talent development and social demand, producing structural imbalances between supply and demand.

To address these tensions, higher education should make further use of AI to improve teaching and assessment, while conducting social experiments in AI and education.

Continue improving teaching models and assessment methods.

Diverse approaches, including blended online and in-person teaching and project-based learning, can enrich students’ learning experiences and improve overall educational effectiveness.

In blended learning, AI can help explore and implement personalised learning pathways and blended course designs, enabling dynamic adjustments to content and difficulty and better allocation of resources.

In project-based learning, AI can recommend project topics that match students’ interests and abilities, intelligently coordinate relevant resources and provide real-time feedback through intelligent project-management tools, effectively developing students’ capacity for innovation.

Assessment should use varied methods to understand and accurately evaluate students’ logical, critical and creative thinking. Alongside traditional tests, these methods can include discussions of open-ended questions, live oral defences and evaluations of project innovation. Intelligent assessment systems can use natural language processing and big-data analysis to examine students’ thinking and problem-solving processes, supporting a comprehensive assessment of their abilities across multiple dimensions.

Deepen educational social experiments for the intelligent era.

Educational social experiments can provide strong evidence and methodological support for exploring how AI is integrated into higher education. By examining changes in educational philosophy, teaching models and learning paradigms, they can reveal the mechanisms through which intelligent technologies affect higher education and accelerate its innovative development.

As AI develops rapidly, universities in China and abroad should strengthen cooperation in using AI for higher education reform and jointly conduct educational social experiments. These should observe effects on educational models, learning efficiency, teaching methods, course design, assessment systems and administration. Analysis of experimental and practical data can investigate AI’s adaptability and effectiveness in different educational settings and cultural contexts, providing tools and examples for higher education institutions worldwide.

What problems may arise, or require particular attention, when AI is used in higher education?

The ethical issues arising from continuing technological innovation generally involve three aspects: unequal rights to technological services, technological risks, and threats arising during use. These are interconnected rather than entirely separate, together shaping the issues that require attention as technology develops and is applied.

The digital divide may intensify educational inequity.

The digital divide is a form of social exclusion, and AI may deepen it and produce new inequities. Universities differ in their digital infrastructure, while teachers and students from different backgrounds also differ in their technological rights and ability to adapt to new technologies.

Unpredictable technological risks are ever-present.

As digital systems become more complex, a gap can emerge between technological methods and their purposes, generating risks that are difficult to anticipate.

1. Data leaks and privacy. Intelligent systems hold extensive information about personal behaviour. This may expose personal information, commercial secrets and even information concerning national security.
2. Opaque algorithms and internal structures. Highly complex algorithms introduce uncertainty in application and may contain bias and discrimination. AI and related technologies can generate results reflecting the value preferences of particular groups.
3. Incorrect information from generative AI, including ChatGPT. Although chatbots have performed well in many business, legal and academic examinations, small changes in a question’s wording can lead to errors or even fabricated answers.

Several potential threats coexist.

Misuse, overuse or avoidance of AI in higher education can threaten individual teachers and students, teacher–student relationships and relationships among students, creating ethical controversies and regulatory difficulties.

1. Misuse. Activities involving intelligent technologies may depart from basic ethical principles and educational requirements, distorting their purpose. Some students, for example, use ChatGPT to write papers or technology to cheat in examinations, causing crises of academic integrity.
2. Overuse. The intelligence and convenience of technology products’ content and services can readily encourage excessive use. Disorganised information overload, stimulating entertainment content, and reliance on outsourcing memory and decision-making affect students’ development. The original interview identifies cumulative harm to physical and mental health, dulled thinking, digital addiction, memory decline and social difficulties among the negative effects. Without a careful response, physical, cognitive and social capacities may face systemic deterioration.
3. Avoidance. Some teachers and students fear, avoid or reject technology. In particular, failing to engage with educational applications of AI may threaten teachers’ sense of agency and the value of their role, increase employment pressures and unemployment risks, and forfeit opportunities for the systemic integration of technology and education.

How should we assess AI’s value for educating university students and transforming education?

AI’s value is multidimensional: it can support personalised learning, enable changes in teachers’ methods, upgrade institutions’ intelligent learning environments and advance innovation in research. Its central value lies in preparing new citizens for the intelligent era, with attention to the following competencies.

1. Lifelong self-directed learning. This will become a core learning competence in the intelligent era. Students should maintain self-efficacy and agency in higher education, continually renewing their knowledge through self-planning, independent decision-making, self-monitoring, self-management and self-evaluation to meet future needs in work and life.
2. Innovation through AI. Using AI creatively requires more than basic technical skills. It involves integrating AI with prior experience to solve complex everyday problems creatively and combining different perspectives to develop innovative solutions.
3. Adaptability across repeated transitions in flexible employment. Students need strong professional skills and interdisciplinary literacy to respond flexibly to changing workplaces, move between roles and responsibilities, meet different occupational requirements and adapt effectively to developments across industries.
4. Handling uncertainty. Responding to social uncertainty requires critical thinking and risk assessment, enabling agile and adaptive judgements and decisions in rapidly changing, complex settings and unfamiliar challenges. Students also need awareness of technological ethics and ethical judgement to use resources responsibly and protect privacy and information security.
5. Living in AI-rich environments. In increasingly close and complex human–AI collaboration, individual agency should remain a priority while human creativity and machines’ ability to handle multiple tasks are combined to achieve the best outcomes.

How should teachers and students respond to AI’s opportunities and challenges?

In the intelligent era, both teachers and students need to adapt to changes in educational philosophy, including fundamental shifts in how they understand knowledge, learning, curricula and teaching.

1. A view of knowledge based on collective creation and sharing. Every individual is a producer and communicator of knowledge.
2. A view of learning based on intelligent connections and knowledge construction. Connected, integrated intelligent learning environments require learners to take a more active role in deep learning.
3. An integrated and open view of curricula. Courses should move from isolated, singular and fully predetermined provision for groups towards integrated, diverse provision responsive to individuals.
4. A view of teaching based on human–AI collaboration. Teachers should coexist, teach and learn with AI while retaining the value of human agency and their capacity to act in their professional role.

Teachers: changing roles and mutual growth through teaching and learning

As a fundamental force in delivering high-quality education, teachers should consciously recognise their changing roles and actively explore the educational benefits of intelligent technologies.

1. Develop awareness of changing roles. With support from intelligent machines, future teachers will inspire and guide students. Alongside maintaining agency and the value of their role, they should provide emotional support and remain flexible in response to students’ learning, interests and character. Teaching should encourage students to express their own ideas and use project-based learning and teamwork to develop practical abilities and creativity.
2. Enable teachers and students to learn from one another. First, teachers should observe and listen to understand students’ real learning needs. Second, they should continually learn about new technologies, set aside assumptions of authority and learn from students about using tools such as ChatGPT to improve efficiency. Third, they should develop their ability to generate questions, exploring open-ended questions with students, challenging established knowledge frameworks and guiding deep learning with AI. Sustained, effective teacher–student–AI collaboration can make teaching and learning mutually empowering.

Students: self-directed learning, real problems and social responsibility

Students should strengthen self-directed learning and the ability to solve complex problems to maintain their competence in the intelligent era.

1. Motivate yourself and strengthen your ability to learn. In a rapidly evolving intelligent era, new technologies and theories continually emerge, making self-directed learning a central driver of personal development. Individuals should maintain self-efficacy and agency, motivate themselves, seek learning resources, set goals and adjust strategies flexibly to meet future needs in work and life.
2. Improve problem-solving through practice in real situations. University students should consciously consider problems from multiple and cross-cultural perspectives, using immersive, participatory and practical problem-solving to develop systems thinking, innovative thinking and critical thinking. They should also make agile, adaptive judgements and decisions in rapidly changing, uncertain, ambiguous and complex environments, using AI efficiently to solve real problems creatively.
3. Take social responsibility in the intelligent era. University students should strengthen awareness of AI ethics and use technology in accordance with relevant requirements. They should also attend to social problems, recognise their responsibilities, consider how AI can help address these problems and actively contribute to a fair, harmonious and sustainable intelligent society.

Source and credits

Original source credited in the archive: Education New Observer (教育新观察) account on Xuexi Qiangguo.

Author: Huang Ronghuai.

Education New Observer, Xuexi Qiangguo reporter: Gu Wenting.

Edited and translated from the Chair website’s archive dated 9 July 2024. The original questions and answers, examples and author’s views are retained; roles and descriptions of technology use reflect the original historical context. Chinese archive

Courses and Resource Centre