Adult learning systems are being called upon to prepare workers for major labour market shifts, including the green transition and the rise of artificial intelligence (AI) in the workplace. While initial education is important, upskilling and reskilling the existing workforce is essential to help individuals and businesses adapt and prepare for these transitions. This chapter provides an introduction to aligning the supply and demand for training for the green and AI transitions.
Training Supply for the Green and AI Transitions
Equipping Workers with the Right Skills
Abstract
Adult learning systems are being called to prepare workers for profound changes in the labour market. Combatting climate change and environmental degradation while harnessing the opportunities presented by greater adoption of artificial intelligence (AI) require a workforce equipped with the right skills. Initial education plays a crucial role in this effort but is not sufficient, as these changes will also require upskilling and reskilling of the existing workforce. Therefore, it is crucial for governments and social partners to invest in upskilling and reskilling programmes to ensure no one is left behind.
Demand for skills associated with the green transition is rising. An analysis based on LinkedIn data shows that the growth in demand for skills for the green transition is already outpacing the growth in supply, indicating a potential shortage of these skills in the near future. Between 2022 and 2023, the presence of green talent in the workforce increased by a median of 12.3% across 48 analysed countries, while job postings requiring at least one skill related to the green transition grew nearly twice as fast, rising by a median of 22.4% (LinkedIn, 2023[1]). In addition to requiring knowledge related to renewable energy and other technologies, green-oriented occupations prioritise process skills, such as critical thinking, monitoring and active learning. Occupations that have more recently emerged as green-driven also require higher levels of these skills than existing green-driven occupations suggesting newer jobs may necessitate even more upskilling (see Figure 1.1).
Many jobs require adapting to the adoption of AI in the workforce. Recent analysis finds that about one in three job vacancies are exposed to AI in some way (Figure 1.2). This means that a significant share of jobs in OECD economies are influenced by AI, even if they do not require specific AI skills. This highlights the importance of training in general AI literacy, to enable workers to use and interact with AI systems. In contrast, only about 1% of high AI exposure job vacancies require specific AI skills (Green, 2024[2]). These jobs are associated with the development and maintenance of AI systems and therefore tend to demand high-level AI skills. Demand for these AI-specific skills, such as machine learning, neural networks and natural language processing, has risen between 2019 and 2022, albeit from a low base (OECD, 2023[3]). Therefore, whilst a large portion of training provision should focus on general AI literacy in order to reach a wider population, courses designed to train specific AI skills remain important as the adoption, use and development of AI in the workforce continue to grow.
Standardised skill level requirements for established and new green-driven occupations
Note: The figure shows the level at which a particular skill is required or needed to perform the occupation. Means have been standardised to a 0 to 100 scale, where greater values imply that a given skill is required at higher levels.
Source: OECD (2024[4]), OECD Employment Outlook 2024, https://doi.org/10.1787/ac8b3538-en.
Share of vacancies with high exposure to artificial intelligence, 2021‑22
Note: Share is defined as the share of vacancies with high AI exposure over all vacancies in the sample by country. Average is an unweighted.
cross-country average. AI exposure is defined by the occupation of each vacancy according to Felten, Raj and Seamans (2021[5]) High-exposure occupations have an exposure measure at least one standard deviation greater than the average.
Investing in skills development for the green and AI transitions can support countries to capitalise on the opportunities presented while mitigating negative impacts on vulnerable groups. The transition towards green technologies and increased AI adoption present substantial opportunities for economic growth, sustainability, and innovation. At the same time, they bring risks of job displacement, exacerbating skill gaps, and widening socio-economic inequalities.
Low-skilled and low-educated adults face a heightened risk of job displacement as a result of the green transition. Previous OECD reports have summarised the literature on the potential labour market impacts of the green transition (OECD, 2023[3]). Current green-driven jobs are disproportionately held by adults with high levels of education and in high-skilled occupation categories, such as environmental scientists and renewable energy engineers. Though future, green-driven jobs may shift toward medium- and low-skilled occupations in activities such as waste management, retrofitting and construction, individuals with lower levels of education are still at the highest risk of displacement due to the green transition.
The risk of job displacement with the adoption of AI remains unclear, but AI creates new tasks and jobs and adults with limited digital skills are at risk. Unlike previous automation technologies, AI has the ability to automate non-routine, cognitive tasks (OECD, 2023[6]). This means that AI will affect not only low- and medium-skilled occupations, as with previous automation technologies, but also high-skilled occupations. Consequently, jobs traditionally held by high-skilled workers could potentially shift to lower-skilled workers. There is some evidence from within workplaces that AI could narrow productivity disparities between high and low-skilled workers, by empowering low-skilled workers to perform similar tasks to high-skilled workers (Brynjolfsson, Li and Raymond, 2023[7]; Georgieff, 2024[8]). While this finding suggests a lower risk of job displacement for low- and medium-skilled workers than with previous automation technologies, it is not clear what the general equilibrium effects might be if entire industries get automated. Irrespective, the impact of AI on tasks and jobs will engender changing skills needs. While companies using AI say they provide training for AI, a lack of skills remains a major barrier to adoption and may pose challenges for adults with limited digital skills, particularly older adults and those with lower overall skill levels, emphasising the need for tailored skills policies.
Efforts to promote resilience through skills policies must prioritise vulnerable individuals, many of whom are less likely to train. Addressing skill gaps and reducing vulnerability through policy action is essential for ensuring inclusive transitions. However, participation rates in adult learning programmes vary significantly across socio-demographic groups, with lower rates observed among those most in need of upskilling and reskilling, such as low-skilled adults, older adults, and people living in rural areas. The disparity between low-skilled and high-skilled adults is striking: while 22% of individuals with tertiary education engage in learning activities each month, this figure drops to only 8% among those with lower secondary education (see Figure 1.3). Additionally, workers in emission-intensive and green-driven occupations, as well as those in jobs at high risk of automation (Figure 1.4), are less likely to participate in training than the average (OECD, 2024[4]; Lassébie and Quintini, 2022[9]). Tailored policies are thus necessary to ensure inclusivity in training opportunities and to effectively equip all segments of society with the skills to navigate the evolving demands.
Share of adults 25‑64 having participated in formal and non-formal learning in the previous 4‑weeks, 2021
Note: Unweighted average for OECD member countries where data is available; contract type and establishment size refer to employees only.
Source: European Labour Force Survey.
Plot of automation index against propensity to train, by occupation
Note: The x-axis represents the automation index computed for detailed occupations (SOC 3 digits). The y-axis represents the share of workers that took part in training in the last four weeks preceding the interview. The sample is restricted to employed individuals in Austria, Belgium, Switzerland, the Czech Republic (Czechia), Germany, Denmark, Estonia, Spain, Finland, France, Greece, Hungary, Ireland, Italy, Iceland, Lithuania, Latvia, Luxembourg, the Netherlands, Norway, Portugal, Sweden, the Slovak Republic, and the United Kingdom. Relative weights were used to construct a crosswalk between ISCO‑08 (three‑digit) and SOC 2010 (three‑digit).
Source: Lassébie and Quintini (2022[9]), “What skills and abilities can automation technologies replicate and what does it mean for workers?: New evidence,” https://doi.org/10.1787/646aad77-en.
The training and skills necessary to prepare adults for these transitions are starting to become clearer. Research indicates that skill requirements for both green-driven and emission-intensive jobs are similar, but targeted retraining in certain skills (especially technical skills, such as installation, maintenance of machines, programming, etc.) is still needed to ensure that workers can successfully transition into green employment (OECD, 2024[4]). The development and maintenance of AI systems necessitates a broad range of skills, including specialised AI skills, basic digital literacy, data science competencies, and complementary cognitive skills (OECD, 2023[6]). Not all workers will require training to develop and maintain AI systems, but most will require skills for adopting and interacting with AI systems, including general AI literacy (Lassébie and Quintini, 2022[9]).
Currently, there is a gap in our understanding about the supply of training provision. Studies about training needs often focus on the demand side – the skills and knowledge that employers are having trouble sourcing – but say little about the supply side – whether and how training providers offer relevant course content. By better understanding the nature of existing training provision, both its content and delivery, we can more accurately assess whether it meets the needs of adult learners in terms of flexibility, accessibility and content. What quantity of courses can adults select from? In what format are they offered – online, in-person, hybrid? How accessible are the courses in terms of duration, but also in terms of the level of pre‑requisites? Finally, are training courses delivering the knowledge and skills necessary to equip workers for changing job demands, including those related to the green transition and greater adoption of AI in the workplace?
This report uses both quantitative and qualitative approaches to shed light on the current supply of training and to assess the capacity of adult learning systems to impart the skills required for the green and AI transitions. To gain a better understanding of existing training initiatives, the OECD distributed two policy questionnaires to identify national and sub-national policies and strategies relating to upskilling or reskilling for the green and AI transitions. The policy questionnaires were circulated to OECD member countries during the spring of 2023. A total of 27 responses were received for the questionnaire on training for the green transition and a total of 21 responses were received for the questionnaire on training for AI (see Annex A for a full list of respondents). The report also provides quantitative insights based on an analysis of training course catalogues in four countries: Australia, Germany, Singapore and the United States. Courses were classified as delivering green or AI content based on whether a list of keywords appeared in their course title or description. The analysis sheds light on the share of courses with green and AI content that are available in these four countries and how accessible and flexible they are across a range of characteristics.
References
[7] Brynjolfsson, E., D. Li and L. Raymond (2023), Generative AI at Work, National Bureau of Economic Research, Cambridge, MA, https://doi.org/10.3386/w31161.
[5] Felten, E., M. Raj and R. Seamans (2021), “Occupational, industry, and geographic exposure to artificial intelligence: A novel dataset and its potential uses”, Strategic Management Journal, Vol. 42/12, pp. 2195-2217, https://doi.org/10.1002/smj.3286.
[8] Georgieff, A. (2024), “Artificial intelligence and wage inequality”, OECD Artificial Intelligence Papers, No. 13, OECD Publishing, Paris, https://doi.org/10.1787/bf98a45c-en.
[2] Green, A. (2024), “Artificial intelligence and the changing demand for skills in the labour market”, OECD Artificial Intelligence Papers, No. 14, OECD Publishing, Paris, https://doi.org/10.1787/88684e36-en.
[9] Lassébie, J. and G. Quintini (2022), “What skills and abilities can automation technologies replicate and what does it mean for workers?: New evidence”, OECD Social, Employment and Migration Working Papers, No. 282, OECD Publishing, Paris, https://doi.org/10.1787/646aad77-en.
[1] LinkedIn (2023), Global Green Skills Report 2023, https://economicgraph.linkedin.com/content/dam/me/economicgraph/en-us/global-green-skills-report/green-skills-report-2023.pdf.
[4] OECD (2024), OECD Employment Outlook 2024: The Net-Zero Transition and the Labour Market, OECD Publishing, Paris, https://doi.org/10.1787/ac8b3538-en.
[6] OECD (2023), OECD Employment Outlook 2023: Artificial Intelligence and the Labour Market, OECD Publishing, Paris, https://doi.org/10.1787/08785bba-en.
[3] OECD (2023), OECD Skills Outlook 2023: Skills for a Resilient Green and Digital Transition, OECD Publishing, Paris, https://doi.org/10.1787/27452f29-en.