For partners
Over the past ten years, artificial intelligence has evolved from the field of theoretical research and engineering applications into one of the key pillars of the global knowledge economy. From 2017 to 2024, the number of English-language university programs containing Artificial Intelligence in the name tripled, from about 150 to more than 450 (according to Stanford's AI Index 2024). In parallel, according to the OECD Skills Outlook, the demand for specialists with competencies in machine learning and data analysis has grown by 28% on average across OECD countries, and in the higher education sector, the share of students choosing AI-related disciplines has tripled over the past ten years.
AI is becoming the new grammar of the modern university — the same mandatory component of scientific and engineering education as the computer science course became in the middle of the 20th century. At the same time, the emphasis is changing: if at the early stage the focus was on algorithms and machine learning, today there is a growing need for specialists in trusted AI, ethics of automated decisions, system integration, and legal regulation of digital models.
Features of studying AI in different countries: how do they do it?
United Kingdom (University of Oxford)
The Oxford School of Artificial Intelligence is based on a research model that combines fundamental approaches and interdisciplinary interaction. The university has a Centre for Doctoral Preparation in AI for the Sciences, where AI is used in chemistry, biomedicine, and climate research: courses include probabilistic modeling, causal inference, interpretive machine learning, and the principles of data ethics. Particular attention is paid to the development of safe systems and the assessment of their impact on society – a direction supported by the UKRI state fund.
Germany (Technical University of Munich, TUM)
The Technical University of Munich demonstrates an engineering and applied paradigm: its Faculty of Computer Science is merged with the Munich Data Science Institute and the Munich School of Robotics and Machine Intelligence. The curricula of master's programs are built around practical tasks: autonomous systems, robotics, signal processing, machine vision. Each student undergoes an industrial internship, and the final project is carried out jointly with a partner from industry. AI is not considered as a separate discipline, but as a universal tool for engineering design.
Netherlands (University of Amsterdam)
Amsterdam's AI program is one of the oldest in continental Europe, operating since the late 1980s. Today, the university is part of the European ELLIS network and specializes in multimodal systems, language models, and AI in media. The courses form the structure of AI and Society, where technical subjects are complemented by the philosophy of technology and digital ethics. The Human-Centered AI master's program involves joint projects of students from the faculties of law, sociology, and computer science — as a result, graduates gain skills in critical analysis of algorithms and their social consequences.
France (ENS Paris-Saclay and INRIA)
The French model is dominated by the mathematical basis. The Master's program in Mathematics, Vision and Learning (MVA) at ENS Paris-Saclay, created with the support of the national institute INRIA, prepares researchers in the fields of optimization, statistical learning, and computer vision. The program is closely connected with scientific laboratories and industrial partners, most graduates continue their studies in graduate school. The French system "nurtures" not so much programming as analytical thinking and the ability to develop new methods for studying models.
USA (Carnegie Mellon University)
Carnegie Mellon is traditionally considered a pioneer in the field of artificial intelligence. The world's first School of Computer Science with a separate faculty of AI has been created here: programs cover robotics, machine learning, cognitive systems, and human-machine interaction. The key feature is the synthesis of engineering and humanitarian approaches: students study the basics of psychology, communication, and interface design. At the master's level, applied areas are actively developing, including automation of production processes and autonomous transport.
Canada (University of Toronto and Vector Institute)
The Canadian model is an interesting mix at the intersection of the university and industrial ecosystems. The University of Toronto is the birthplace of deep learning: Geoffrey Hinton and his school of neural network research work here. In 2017, the Vector Institute for Artificial Intelligence was established, which, together with the university, develops curricula and finances scholarships for undergraduates. Students are involved in real-world research projects, and courses include both mathematics and programming, as well as analysis of the social impact of AI. Canada is committed to sustainable development and responsible technology – this is reflected even in the curriculum standards.
China (Tsinghua University)
Tsinghua University in Beijing has become the center of the national AI strategy. In 2018, the Institute for Artificial Intelligence was opened under it, which united the laboratories of computer science, mathematics, microelectronics, and philosophy. The programs are focused on the development of scale models and specialized AI chips, students study not only machine learning, but also the design of computing architectures, national security issues, and data ethics. The studying is accompanied by participation in government projects and internships in the technology sector.
A similar model is used in Singapore, Indonesia and South Africa, adjusted for local capabilities.
Japan (University of Tokyo and RIKEN AIP)
The University of Tokyo is developing fundamental and applied AI in collaboration with the RIKEN Center for Advanced Intelligence Project (AIP). Here, the emphasis is shifted to humanitarian and social aspects: explainability of solutions, human-robot interaction, AI for medicine and materials science. The program emphasizes the idea of "ethical engineering" – the development of technology while preserving the human dimension. The educational process is based on the principles of project-based learning: students participate in international competitions, conduct joint research with laboratories and industrial companies.
Comparing regional priorities
Regional educational approaches to AI are determined not only by academic traditions, but also by the socio-economic priorities of states.
- The European model is based on an ethical and legal context and institutional regulation. The programs of EU universities are subject to the pan-European framework of Ethical AI and the standards of transparent algorithms. A course on "responsible AI" is becoming a mandatory element even in engineering master's programs. Universities actively cooperate with supranational structures — the EU AI Act, OECD AI Principles, and national agencies for digital regulation.
- The North American model shows a different emphasis, entrepreneurial and exploratory. Universities in the United States and Canada are building close ties with the industry by creating collaborative labs, venture incubators, and applied master's programs. Here, AI education is not separated from the innovation ecosystem: students are involved in startup projects, and the faculty participates in corporate research.
- The Asian model, on the contrary, has a state-institutional character. In China, Singapore, India, and Japan, the development of AI education is included in national strategies for digitalization and technological sovereignty, universities act as executors of state plans, and the preparing of specialists is correlated with the priorities of industrial policy. Asian programs are distinguished by their scale: specialized AI colleges, national laboratories, and centralized data exchange platforms are being created. The main benchmark is the speed of implementation and the practical return of technologies.
What will AI education be like in 5 years?
In the foreseeable future, education in the field of artificial intelligence will develop towards the breadth of coverage and diversification of formats. Three interrelated trends determine its future configuration:
- Massification of AI Competence
If in the past the preparing of specialists was limited to the faculties of computer science, today elements of AI are being integrated into the programs of economics, law, biomedicine, architecture and public administration.
- Development of professional master's and micro-qualifications
Against the backdrop of a shortage of teachers and the rapid obsolescence of technology, the importance of flexible forms of education is growing. Increasingly, such programs are being implemented in partnership with industry, ensuring that they are relevant and practical.
- Institutionalization of responsibility
In the coming years, the regulatory environment will become one of the main factors in educational planning. The adoption of the EU AI Act, OECD recommendations and similar initiatives in Asia require the inclusion of modules on ethics, safety and regulation of AI in the curriculum.
The world map of educational practices demonstrates a variety of approaches, but a single vector is a combination of scientific depth, engineering accuracy and social responsibility. Europe seeks to institutionalize the ethical and legal side of technology, North America seeks to turn AI into an engine of innovation and entrepreneurship, and Asia seeks to build it into national development strategies. As a result, we can expect the emergence of a new type of specialist: a researcher and practitioner who is able to think interdisciplinarily, to see AI not only as a set of algorithms, but also as a social phenomenon. Universities are becoming a space where a culture of responsible use of intelligent systems is being formed, a culture on which the humanistic future of digital civilization largely depends.
