Cedefop presented new evidence on the demand for artificial intelligence (AI) skills, as revealed in the job vacancies of EU employers, and its research on automation risk and the responsiveness of vocational education and training (VET) systems to the future of work challenges, at an international conference on AI and education in Beijing on 16 to 18 May.  

The conference was organised by UNESCO, the Ministry of Education of the People’s Republic of China, the Chinese National Commission for UNESCO and the Beijing Municipal People’s Government. Over 230 representatives from 105 member states took part, including more than 60 ministers, deputy ministers or permanent secretaries. Around 100 speakers and participants from UN agencies, international organisations, academia, civil society and the private sector discussed emerging policies and strategies for leveraging AI’s possibilities for reshaping education provision and allowing progress towards meeting the UN’s strategic development goals (SDGs) of the education 2030 agenda. The conference gave opportunities to participants to network with top-level Chinese experts in the AI area and leading Chinese companies in the field of AI in education.

It was recognised that AI can help countries overcome major challenges in achieving the SDGs, such as reducing barriers to education access, automating management processes, analysing learning patterns and optimising learning processes with a view to improving learning outcomes.

Employers’ needs for AI skills

Cedefop expert Konstantinos Pouliakas presented evidence from the agency’s digitalisation and future of work project at a session on anticipation and development of skills needed for work and life in an AI era. The session focused on the capacities of education and training systems to proactively respond to changes in labour markets, including: what occupations are at risk with the advent of AI, and what new occupations are being created; how education and training systems can anticipate these changes to equip the existing workforce and prepare new generations with job skills to succeed in the AI era; how the private sector can be involved in the governance, financing and management of technical and vocational education and training programmes.

Cedefop showed evidence on the extent to which automation is likely to pose a risk for occupations and sectors in EU labour markets. Based on unique data on the skill needs of EU jobs from the European skills and jobs survey, Mr Pouliakas argued that about 14% of EU adult employees’ jobs face a very high risk of automation. Around 40% of EU jobs are likely to see a significant part of their skill needs and tasks transformed, highlighting the critical importance of investing in employees’ continuing skill development.

The presentation also showcased first efforts by Cedefop to extract information on the demand for AI skills as expressed by employers in 11 EU countries. Using novel information from Cedefop’s prototype and newly released skills online vacancy analysis tool for Europe (Skills OVATE), evidence was presented that demand for AI skills (e.g. machine learning, deep learning, image recognition, natural language processing etc.) is still concentrated in specialised ICT fields, such as systems analysts. But it is also gaining traction among database and network professionals, electrical engineers and technicians and metal working machine tools operators.

The tool highlights that AI skills are part and parcel of a whole array of different skill needs, most notably soft skills like adaptability to change, foreign languages for international careers, project management and teamwork, even among highly specialised AI occupations.

VET in an AI era

This evidence implies that a truly AI-proof education is one that seeks to employ an interdisciplinary approach in which data and technological/digital skills are complemented by other key competences. A recent data collection by Cedefop’s ReferNet network on the responsiveness of EU VET systems to the future of work technologies underlines that many EU countries are seeking to reform their VET programmes and curricula towards that direction, combining science, technology and engineering with arts and media design courses.

As part of its digitalisation and future of work project, Cedefop continues to investigate the implications that AI will have on EU labour markets, including investigation of how algorithmic management of workers employed in the platform economy is affecting their skills development and mismatch (CrowdLEARN study).

Cedefop has launched the 2nd European skills and jobs survey to collect comparative EU-wide information enabling investigation of the impact of technological change and digitalisation (including automation) on workers’ skill mismatch/obsolescence and their readiness to adapt to changing skill needs via remedial learning practices. Cedefop will also continue to collect evidence on VET policies and practices adopted by EU Member States to ensure that they are digital frontrunners in the technology race and that all EU citizens can enjoy the marked benefits that will be made possible due to rapid technical advances in the AI era.