Skills empower workers in the AI revolution: A summary of Cedefop’s first AI skills survey

Artificial intelligence (AI) is the next general-purpose technology rapidly transforming labour markets, industries and workforce skills, creating both new opportunities and significant challenges. The Cedefop AI Skills Survey 2024 shows that more than a quarter of the European adult workforce is already experimenting with the use of AI at work and, with the demand for AI-related skills growing, emphasises that upskilling, reskilling and investing in AI literacy will be crucial drivers of a human-centred AI revolution.

AI jobs and skills

Cedefop’s digital transition policy scenario, based on the European skills forecast framework, examines the impact of automation and AI on employment. While rapid AI deployment could reduce employment by 5%, potentially displacing up to 7.5 million workers, the full effects remain uncertain due to complex market adjustments.

The research suggests AI’s job-destruction impact is lower than its job-augmentation potential. However, challenges in data representation hinder accurate mapping of AI adoption in European workplaces. Additionally, focusing solely on technical AI skills, e.g., machine learning, NLP, overlooks broader societal and economic implications.

The rise of AI in the workplace

Over the past decade, the demand for AI skills has surged, and many large European companies already implementing AI technologies, however, AI adoption varies across countries and sectors, from over one in four in the information and communication sector to less than one in thirty firms in the construction sector. Additionally, Western European countries, such as Belgium, Germany, France, and Luxembourg, lead AI adoption, while Southern and Eastern European nations have lower usage rates, raising concerns that a two-tier Europe may emerge, and deepening the already existing digital divide.

AI’s influence on labour markets is profound but complex, and while initial fears of mass unemployment are subsiding, AI is primarily transforming job roles rather than eliminating them outright. However, specific groups, particularly workers with lower education levels, in middle-skilled roles, and in contexts where mostly computerised machinery is deployed face higher risks of job displacement, and will need upskilling and reskilling opportunities to remain competitive in the evolving job market.

AI’s impact on jobs and skills

Key findings in the AI Skills Survey include:

• 15% of European employees fear job loss due to AI automation within the next five years, with the highest concerns in Greece (24%), Poland (19%), and Slovakia (19%).
• More than one in four (28%) European workers already interact with AI at work, either directly or through AI-powered systems used by colleagues.
• AI can boost efficiency: 67% of workers using AI report that it helps them complete tasks faster, while 41% say it enables them to take on new responsibilities.
• AI-driven task transformation is highest in finance, retail, and manufacturing, whereas education, healthcare, and transport experience lower automation risks.

Despite these transformations, the report highlights a significant gap in AI literacy with many workers lacking the necessary AI skills, and only 15% of workers actively participating in AI training programs. Older employees, women, and those in precarious employment contracts are least likely to engage in upskilling programs, raising concerns about unequal access to AI education.

While AI can and does have the potential to improve efficiency, the survey found that over half of AI users (54%) feel that it has not made them better at their jobs, mainly due to low AI literacy levels and insufficient training opportunities. These findings may reflect that the survey took place at a time when the use of AI technology in workplaces is only just taking off, but they are cause for concern, highlighting that without investing in people and their skills, AI’s benefits may be unevenly distributed, exacerbating inequalities rather than fostering inclusive growth, as the recent Draghi report on the future of European competitiveness has also pointed out.

Competing in the AI race means investing in skills

The survey highlights that with six in ten employees expected to experience AI-driven task transformation, investing in AI literacy, worker protections, and ethical AI policies will be crucial for building a resilient and inclusive digital economy.

Promoting inclusive AI literacy

In occupations at all levels, including highly skilled ones, workers need skilling to cope with the AI-driven transformation of their jobs. However, the survey highlighted that a substantial share of workers (44%) expresses concern that they will not receive adequate training to work with AI technology.

Therefore, embedding an AI competence framework into initial education and training is vital for adequately preparing new learners for the future AI-driven job market, and it should also be a core priority in adult learning and continuing vocational training, as AI literacy in the European adult workforce is low. Around 40-60% of adult workers surveyed have a poor understanding of what AI technology is, how it operates and what its likely consequences for society are.

Targeted vocational education and training initiatives should also be designed in ways that overcome gender and age segregation in the use of AI technologies. And even though highly qualified workers are more inclined to use AI technology in their job, they also need access to AI skilling. The need to upskill in AI spans across educational levels.

Empowering social dialogue

Empowering social dialogue and a culture of workplace trust is a critical factor in ensuring that AI diffusion will not result in worker resistance, quiet quitting or other behavioural obstacles, e.g., lack of awareness, reservations against AI or lack of proactivity.

The survey highlights the need for organisational incentives and support that empowers workers and a shift away from approaches that introduce or magnify the fear of automation. It also finds that AI take-up and training are higher in organisations where a trade union or other worker representation body is in place.