The Dutch labor market is strong and evolving: labor force participation is notably high, and unemployment levels are historically low. But ongoing trends, including an aging population and declining productivity growth, are putting the labor market under pressure. Recently, McKinsey projected that labor market tightness could triple by 2030 if the Netherlands maintains current levels of growth in GDP (1.6 percent CAGR) and productivity (0.4 percent CAGR).
While traditional automation solutions to increase labor productivity, such as document processing systems, have played a significant role in addressing this challenge, most have been limited to processing structured data. Generative AI (gen AI), however, unlocks a new domain of automation with its ability to process unstructured data such as natural language and images. It therefore broadens the spectrum of automation potential to more occupations, including knowledge work and customer service, and holds the potential to boost productivity and economic growth for the Netherlands.
Following the base case scenario of automation of current activities (“slower scenario”) from a recent McKinsey Global Institute (MGI) report on the future of work in Europe, we estimate that automation will likely affect 15 percent of total full-time-equivalent work hours in the Netherlands by 2030—a 50 percent increase over a scenario without gen AI. This represents a significant opportunity for Dutch businesses to tackle the challenges posed by the evolving and tight labor market, especially in areas in which increased automation can help relieve labor shortages. However, a wide deployment of gen AI is expected to require extensive re- and upskilling programs, including for jobs that were previously unaffected by automation. It may also require regulation to mitigate its potential risks, such as data privacy, intellectual property risk, and fairness.
In this article, we focus on how gen AI can influence the future of work in the Netherlands across sectors and organizations. We discuss how expected adoption speed varies by sector and by sector composition in terms of shares of small and medium-size enterprises (SMEs), independents, and corporations. We also explore actions public and private leaders could take to accelerate gen AI adoption in the Netherlands. This perspective takes a longer-term view of the singular effects that might be directly brought about by gen AI; it does not forecast aggregated employment effects that can be brought about by the business cycle in the short term.
Gen AI creates new opportunities spanning sectors and organizations
The introduction of gen AI and the public breakthrough of OpenAI’s ChatGPT in late 2022 have accelerated the automation potential of activities that involve communication, documentation, or interaction with people. Gen AI—in particular, applications that are open or that use publicly available data—has also become available to a wider group of organizations. The range of work activities suitable for automation has expanded to include those requiring subject matter expertise, interpersonal interaction, and creativity. Consequently, the timeline for automation adoption could accelerate significantly, reaching previously unaffected jobs, such as those in professional services.
Our modeling suggests that even in the “slower scenario,” adoption of gen AI could reduce the total number of hours needed to perform current workforce activities by 15 percent (Exhibit 1; see sidebar, “Methodology”). Of course, the exact automation path for gen AI is subject to considerable uncertainty. The extent to which this potential is realized will depend on the ability of Dutch companies to innovate, the capability of workers to re- and upskill, and the support of policy makers.
In the rest of this section, we dig into the precise factors affecting uptake as well as the sectors in which gen AI has the most promise for addressing labor market challenges.
Impact from gen AI will depend on sector composition
While our models reveal potential for gen AI to relieve labor market tensions, we also acknowledge the challenges involved in gen AI reaching its full potential in the Netherlands.
Not all Dutch businesses are preparing to adopt analytical AI (that is, AI methods preceding generative AI). AWVN (General Employers Association of the Netherlands) reported that 40 percent of Dutch companies are not yet using AI in their businesses because of a lack of knowledge, safety and privacy concerns, or perceived irrelevance. This finding illustrates the types of challenges the Netherlands may face in rapidly adopting gen AI technologies.
The speed of gen AI adoption at a country level is determined by multiple factors, including the economic maturity of a country, overall sector readiness to embrace new technologies, and crucially, sector composition in terms of shares of SMEs, independents, and corporations (Exhibit 2). A relatively high percentage of workers in the Netherlands (about 65 percent) are employed by SMEs, compared with 57 percent in Germany, 54 percent in the United Kingdom, 52 percent in France, and 46 percent in the United States. Sectors dominated by small companies might be slower to embrace new automation opportunities. For example, when it came to adoption of digital sales technologies, in 2019, the top 10 percent of largest companies captured 60 to 95 percent of digital revenues. This is expected to affect the rate and speed at which gen AI might accelerate automation in the Netherlands.
Large companies in the Netherlands are adopting analytical AI technologies faster than smaller companies. Centraal Bureau voor de Statistiek (CBS) reported in 2020 that 48 percent of companies with 500 or more employees were using one or more AI technologies, compared with only 8 to 13 percent for companies with ten to 50 employees.
As with the adoption of analytical AI, the smaller-company archetype could capture impact from gen AI more slowly in the next five years for the following three reasons:
Limited investment capacity. Smaller companies usually have lower investment capacity, constraining their ability to acquire new capabilities, tools, or resources. For example, Dutch SMEs invest about 1 percent of profits in R&D compared with about 5 percent for larger companies. They might lack the scale to manage these solutions systematically, such as keeping marketing content generator inputs up to date. However, many gen AI solutions—especially those integrated in existing software packages such as Adobe and Microsoft Office—are already available for many SMEs.
Lower up- and reskilling capabilities. Introducing gen AI tools into existing operating models requires significant change management of IT-related processes, including upskilling and reskilling programs to help employees use gen AI tools effectively. Smaller companies often lack the scale to benefit from designing and running such programs and are half as likely to provide formal upskilling programs compared with large corporates. Rather than formal in-person training programs, these organizations might rely more on digital and self-organized training, which could provide a less effective learning environment.
Less robust existing data and infrastructure. The low-complexity technological landscapes of smaller companies typically do not justify extensive investment. Consequently, the technology is generally less mature, and existing data and infrastructure are generally less available. SMEs may, for example, use customer databases manually and not directly link databases to marketing and sales systems. All of this limits the pace at which gen AI solutions can be integrated with existing systems and data.
Challenges to automation adoption are not equal for all sectors in the Netherlands, because the share of SMEs varies widely across sectors (Exhibit 3). For example, the agriculture sector could potentially benefit from using analytical AI and gen AI to improve efficiency for farms and agriculture companies, such as by improving on-farm decision making with camera images. However, the high percentage of independent and smaller companies in the sector along with factors such as plot size or specific legislation has thus far led to lower adoption of automation. In fact, McKinsey research shows that only 33 percent of Dutch farmers, who farm on smaller plots than farmers in neighboring countries, use at least one agricultural technology, compared with 45 percent in Germany and 51 percent in France.
Additionally, we expect that the public sector will experience slower adoption of gen AI. While most subsectors in the public sector have a high share of enterprises with more than 500 employees (about 50 to 80 percent in government, healthcare, and education) and may have more scale to enable gen AI upskilling, public acceptance and different regulatory requirements can potentially slow the adoption of gen AI in this domain. McKinsey research also shows that the overall investment in analytical and gen AI is lower in public sectors than in other sectors.
Gen AI holds the greatest potential to address Dutch labor market challenges in five sectors
Although various interventions can address labor market challenges, our models show that gen AI can be a major productivity booster, particularly in a handful of sectors (Exhibit 4). A previous McKinsey report, Netherlands advanced: Building a future labor market that works, estimated that roughly one-third of the necessary productivity improvements to address labor market tightness in the Netherlands can be achieved through automation powered by gen AI.
While gen AI is a game changer in speeding up automation for some sectors, implementing this technology may face delays due to sector nature and composition. Sectors may require varying levels of innovation orchestration and support to see positive results.
This orchestration could include four elements. The first, raising awareness, could involve helping sectors understand relevant high-impact gen AI use cases and their feasibility. Next, technology such as cost-effective, scalable, and secure gen AI solutions should be available, along with support for AI literacy training. Third, support to establish pilots and measure impact on labor tightness could serve as proof points for broader adoption and scaling. Last, cross-sector partnerships could be established to scale impact, pool funding, and share knowledge—for instance, across public sector, industry, and educational institutions. Below, we explore five sectors with high potential for gen AI automation impact: finance and insurance; IT; administrative and support and government; educational services; and professional, scientific, and technical services.
Two sectors likely to self-propel adoption of gen AI
The finance and insurance sector and the IT sector have the highest potential for gen AI automation and some of the highest labor market tension. Gen AI innovation in these sectors will enhance their productivity. This, combined with their higher share of large corporations and large SMEs (approximately 50 percent of all businesses), provides business incentives and scale to encourage adoption without significant orchestration.
The finance and insurance sector is expected to adopt gen AI because of strong commercial incentives to substantially automate operations and address labor shortages. This sector stands to benefit significantly from gen AI tools, particularly in automating daily tasks, and its high level of digitalization can facilitate the integration of gen AI into primary processes and technology development. For example, ING partnered with McKinsey to develop and deploy a gen AI–powered chatbot. Gen AI can also speed up credit and underwriting applications by automating document and image processing and generating credit memos.
The IT sector can benefit greatly from gen AI’s ability to automate and accelerate software development processes. Gen AI–powered coding assistance, automated testing, and intelligent architecture-design recommendations help IT companies speed up feature delivery and reduce time to market. AFAS, for example, has developed an API integration with OpenAI, giving companies access to advanced AI models for generating business information. This helps automate tasks such as reporting, business data analysis, and document creation. Gen AI can help minimize incident and problem resolution times by creating tighter feedback loops, linking user behaviors to feature changes, and collecting real-time feedback. Delft University of Technology’s AI for Software Engineering Lab (AI4SE) explores these opportunities.
Orchestration can accelerate gen AI adoption in three additional sectors
Our models suggest that most sectors would benefit from orchestration to accelerate gen AI adoption, including the public sector and sectors with a higher share of smaller companies. Here we explore three sectors that will likely require orchestration for gen AI adoption and where the potential for gen AI impact is highest—including the administrative and support and government sector, the educational services sector, and the professional, scientific, and technical services sector.
In the administrative and support and government sector, many traditional tasks are highly suited for automation using gen AI, such as case handling and call center operation. In educational services, early gen AI applications already focus on adaptive tutoring and service chatbots. These tools can enhance learning experiences and alleviate labor shortages by supplementing teacher-led instruction. However, both of these subsectors are in the public sector and are expected to adopt gen AI more slowly, as discussed above. In some cases, however, the government has played a leading role in accelerating adoption to enhance service delivery and support. For example, in the administrative and support and government sector, Rijksdienst voor Ondernemend Nederland (RVO) launched AskSenna, an AI-driven tool designed to assist start-ups and early-stage companies by providing instant answers to regulatory and business-related queries.
In the professional, scientific, and technical services sector, including consulting, solutions are being developed to use gen AI for advanced search, synthesis tasks, and virtual coaching. However, given the high percentage of independents and SMEs in the sector (80 percent), we expect slower holistic adoption of gen AI solutions and therefore a slower impact on labor tightness. This sector may benefit from orchestration that is particularly targeted to help smaller enterprises understand the specific applications of gen AI relevant to their business, learn approaches to successfully implement those applications, and improve AI literacy.
Gen AI will create new roles and occupational categories
As gen AI becomes increasingly integrated into various sectors, the demand for specialized skills and expertise in AI is expected to grow. For example, we expect increased demand in three occupational categories in the Dutch labor market:
Gen AI practitioners. Gen AI specialists—including prompt and agent engineers or AI content auditors—form a new subexpertise within the AI playing field. Globally, these roles have grown rapidly across sectors that implement gen AI in their daily practice, especially in IT functions. Additionally, the surge in gen AI will drive demand for related software and data engineering support. For example, the Dutch company Weaviate helps companies structure their data to facilitate the development of gen AI use cases.
Gen AI researchers. Gen AI has created new opportunities in research positions, within both academia and enterprises. For instance, Philips is developing gen AI applications to improve clinical decisions, diagnosis, and workflow. The Dutch start-up Cradle uses gen AI to predict protein properties that could accelerate vaccine development.
Semiconductor, software, and other engineers. As gen AI continues to grow, semiconductor-related jobs that enable this technology will likewise expand, including semiconductor engineers to provide computing power, software engineers to build front-end solutions, and a wide range of other roles. This presents an opportunity for the Dutch semiconductor industry, and companies such as ASML, ASM, Besi, and NXP Semiconductors are positioned to grow substantially because of the expected increase in semiconductor demand driven by the growth of end applications including gen AI. For example, in April 2024, ASML and the local municipality Eindhoven signed a letter of intent to expand ASML’s facilities to accommodate an additional 20,000 employees—a 50 percent increase in growth, which could be partly driven by gen AI.
We have previously emphasized the importance of developing soft and hard skills to keep pace and enable career advancement. This ongoing development can enhance productivity in current positions and prepare individuals for the high-demand jobs created by and for gen AI.
Three moves could accelerate gen AI adoption and manage its effect on the Dutch workforce
Three actions by public and private stakeholders could accelerate gen AI adoption in the Netherlands and will likely have a positive effect on the Dutch workforce.
Preparing for granular upskilling and reskilling needs
Increased automation requires upskilling workers to use new gen AI tools—monitoring service chatbots or using copilots to write marketing content, for example. Companies will need to develop training programs as these solutions are implemented. And because demand for some professions may decline, workers may need to reskill, sometimes across sectors. This would necessitate greater orchestration. The public sector could facilitate this process—for example, as Techniek Nederland has been doing since 2013 by reskilling individuals from various backgrounds to become installation engineers.
Granular insights such as regional job gain and loss analysis are crucial to understand a company’s reskilling needs and make informed decisions. To assist Dutch businesses in identifying their automation potential and its workforce impact, public sector agencies and businesses could consider developing tools that map different job types to their expected automation potential. Such tools could provide businesses and local governments with valuable insights and guidance for transitioning to gen AI–driven processes, helping them make informed decisions and support employees throughout their journey.
Orchestrating sectors with high degrees of SMEs and labor tightness
In private sectors such as the professional, scientific, and technical services sector, in which both automation potential through gen AI and the proportion of SMEs are high, orchestration can accelerate gen AI implementation. Larger tech corporations, universities, public sector agencies such as UWV, and sectoral employer organizations could potentially facilitate this.
Furthermore, Dutch companies of all sizes have opportunities to engage in the gen AI–based automation market. For example, banks could collaborate with AI or software companies to create specific propositions or loans for gen AI development and help disseminate these throughout the sector. The Nederlandse AI Coalitie (NL AIC), a public–private partnership, aims to promote the adoption and ethical use of AI technologies across sectors.
Making bold investments to lead the Netherlands’ gen AI transition
Strategic investments by both public and private organizations could accelerate the adoption of gen AI and alleviate labor shortages in the Netherlands. Such funding for long-term innovation will be crucial, especially in sectors that are strategically important to the economy, such as manufacturing (including the semiconductor industry), healthcare, construction, and education. For example, NFI, SURF, and TNO have received €13.5 million to develop a Dutch gen AI model that could accelerate development and adoption across sectors.
The Netherlands already hosts a few AI funds, such as NL AIC for responsible AI, AI growth fund AiNed, the Nationaal Onderwijslab AI (National Education Lab AI) established by Radboud University for AI in education, and the Innovation Center for Artificial Intelligence. In the private sector, many large corporations are investing significantly in developing innovative technology. For example, ASML and Philips launched DeepTechXL, a private investment fund to finance and guide high-tech start-ups.
By embracing collaboration on cutting-edge technology opportunities such as gen AI, the Netherlands can position itself to build an increasingly thriving business ecosystem. A proactive approach can accelerate innovation and economic growth as well as ensure the workforce is well prepared to adapt to the changing landscape and flourish in a future shaped by AI.
