AI offers opportunities for the Netherlands, but success is not a foregone conclusion

Whether AI will actually make the Dutch economy more productive depends on three factors: 1) businesses must have access to robust and reliable AI technology, 2) they must be able to productively translate this technology into widely adopted applications, and 3) they must be given the scope to grow. 

AI can have a significant impact on economic growth and inflation, which is why it is important for central banks. DNB’s research focuses on productivity, employment and inflation, and on the conditions under which the Dutch economy can benefit from AI. AI also poses risks to society and the financial system. DNB’s present Analysis does not address the societal questions raised by AI, but DNB researchers have previously examined the risks to the Dutch economy and pension funds posed by a potential fall in the price of AI shares (article in Dutch). 

The key to success lies in the application of AI

AI can help businesses and workers, take over tasks and enable new products and services. Our research shows that the macroeconomic benefits for the Netherlands depend primarily on how businesses use AI and roll out successful applications more widely, as using new technology does not automatically lead to higher productivity. 

To improve their productivity, businesses will have to reorganise their work processes and employees will have to learn new skills. Furthermore, businesses must make good use of the time saved by AI. For instance, they could use it to improve product quality, develop new products, or boost production. In this way, the benefits offered by AI could have a wide-ranging impact on the economy.  

The Netherlands is in a strong position, but there is work to be done

The Netherlands is in a strong position to benefit from the use of AI, with a robust digital infrastructure, a highly skilled workforce and a large number of knowledge-intensive service sectors. Dutch businesses are among the European frontrunners in the use of AI. However, the use of AI is not growing at the same rate across the board. Large businesses and sectors that make extensive use of digital technology are leading the way. Smaller businesses use AI less frequently, as do businesses in the construction, transport and hospitality sectors.

However, it is precisely smaller businesses that can benefit from AI. Using software and cloud services, they can gain access to specialist knowledge that was previously too expensive, without having to develop AI models or their own technical infrastructure. However, there is still work to be done if AI is to be used effectively: business owners need to choose suitable applications, train staff and adapt work processes. Smaller businesses, in particular, do not always have sufficient time or in-house expertise to do so. Industry organisations and knowledge networks can help them by sharing their experiences with AI and offering practical advice. 

Advanced AI models must remain accessible

The Netherlands does not need to develop the most advanced AI models itself in order to benefit from this technology. However, businesses must retain reliable and affordable access to AI models and to the necessary data and computing power. They must also be able to switch easily between providers, so that they are not tied to a single supplier. 

Advanced AI models and the associated infrastructure are developed only in a limited number of countries and by major technology firms. European cooperation can improve Dutch businesses’ access to computing power, cloud and data infrastructure, and offer them more alternatives. Europe does not need to be completely self-sufficient, however.  The main point is that businesses must maintain reliable access and not fall victim to vendor lock-in. 

Work is changing, but the impact on jobs remains to be seen

AI can take over tasks, help workers to be more productive and enable new products and services, and that is why AI does not necessarily have to lead to job losses. It is uncertain how the number of jobs will develop. For many people, the nature of their work may change. Some will be able to adapt within their current role, while others will have to find a new role or move to a different sector.

How work will change depends to a large extent on the choices that businesses make and how they involve their employees in the process. We conclude that employers should involve their employees at an early stage in decisions regarding the use of AI and the skills required in the process. Employers’ and employees’ organisations can set out agreements in their collective bargaining regarding retraining and further professional development, as well as changes to job profiles. This is in line with the advice of the Social and Economic Council of the Netherlands (SER). 

Before they actually lose their job, employees should be facilitated in making the transition to a different role, for example by offering training, career guidance and support in finding a new job.

Successful businesses must be given the scope to grow

Some businesses will be better equipped to use AI than others. The economy benefits more when successful businesses are able to grow and attract more staff and funding. To achieve this, new businesses must be able to start up and existing businesses must be able to grow, but businesses must also be able to downsize or cease trading if they are not sufficiently successful. A previous DNB study shows that relatively few businesses are set up or wound up in the Netherlands. Furthermore, staff, equipment and other resources are not always deployed at those businesses where they would be most effective. 

Policy must be designed to ensure that labour and capital are not unnecessarily tied up in less productive activities. Among other things, this requires a critical review of regulations that may hinder the establishment, growth, downsizing or closure of businesses. At the European level, too, successful businesses must be given more scope to grow. The Netherlands must therefore carefully weigh up additional national requirements for AI against the importance of a well-functioning European market. In addition, a more efficient European capital market could offer businesses more opportunities to finance their growth. Successful AI applications can then be adopted more widely across the economy.