The 2023 AI Skills Needs Analysis report sheds light on the most needed AI roles and skills in Europe. The conclusions are drawn based on the study of job vacancies powered by artificial intelligence, labour market survey reports, available research on AI roles and skills, learning programmes, and databases. We also collected experts’ input on the current and future of AI roles, skills, and provision.

This AI Skills Needs Analysis is a starting point to design a coherent AI Skills Strategy for Europe with skills intelligence on the current and future skills demand in the field of AI. It will also feed the development of the AI Skills Strategy with information on the current and future demand for AI roles and information on the current supply of AI learning offers. This needs analysis also fulfils the first purpose of a Blueprint project: gathering skills intelligence — with the results to be fed into the ESCO classification of occupations and skills & competences.

Main conclusions

Roles with Need for AI Skills

• The AI practitioners’ roles that are needed most are data scientists, data engineers and especially machine learning engineers including NLP engineers and computer vision engineers.
• An emerging role that requires urgent attention is prompt engineer.
• AI management & support roles are also emerging with the most foreseen need for AI strategists, AI ethics officers and AI quality controller.
• Organisational decision-makers like business leaders and middle management, and policymakers all need basic AI knowledge and skills.
• Policy- and decision-makers rely on AI advisors that combine deeper AI knowledge and skills with expertise in policy or business.
• These findings can feed the ESCO occupational roles by suggesting to add (or amend) roles.

Skills Needs of AI Professionals, Policy-makers & Decision-makers

• Each AI professional role (e.g., data scientist, machine learning engineer) has its own specific set of technical skills that can also be different based on the specific context of the AI professional.
• Each AI professional role also needs a set of skills on transversal topics (e.g., ethics, security), soft skills (e.g., problem-solving), and skills on functioning in organisations.
• Policy- and decision-makers need basic AI knowledge and skills including basis terminology & practice, AI ethics, and law & regulations.
• AI advisors need AI advisory skills covering, e.g., AI risk management, AI compliance, AI strategy, and implementing AI.  
• Both Basic AI knowledge & skills and AI advisory skills need urgent attention for Europe to move forward in the field of AI.
• These findings can feed the ESCO Skills & competences by suggesting to add (or amend) skills.

Supply of AI Learning Programmes & How to Close the Skill Gaps

• There are learning programmes for common AI professional roles although not enough people are getting educated or trained in these programmes at the moment.
• There is hardly any supply related to emerging demand like prompt engineering, although those skills are needed urgently. Programmes on these skills need to be agile and modular so they can be updated fast, meeting the quickly changing needs.
• Full educational profiles will need to be developed for AI professional roles that include soft skills, transversal skills and skills-related functioning in organisations.
• Educational profiles for policy- and decision-makers will need to focus on AI-related knowledge and skills in line with the skills needs mentioned above.
• There is an urgent need for AI upskilling programmes. These programmes are a short programme on basic AI knowledge for policy- & decision-makers, a programme on AI advisory skills, and a longer AI practitioner programme on prompt engineering.
• These urgent upskilling programmes will be designed, developed and implemented before the end of the year.