Qualified knowledge yielded by research and innovation is essential to fully understand the transformative potential and far-reaching consequences of AI. It is also an important prerequisite for successful formulation and execution of appropriate governance structures, policy responses, and interventions. As has been presented in this discussion paper, national governments in the Nordics and the EU through its framework programmes have committed significant resources to research and innovation. However, the response is arguably still in its infancy and the national AI strategies in the Nordic countries point towards the importance of research and innovation, laying a foundation for long term investments. In 2019, the OECD published the policy framework “The OECD AI principles” encouraging governments to work together in international and regional fora to foster international, cross-sectoral and multi-stakeholder initiatives to garner long-term expertise on AI.
“International research collaboration on AI has the potential to accelerate scientific progress by combining knowledge, resources, and expertise from different countries and cultures. It can help address and resolve ethical, social, and political challenges associated with the global implementation of AI technologies on a broader, international scale.” (OECD 2023, p 60.)
In this context, the NordForsk administration believes that a regional funding scheme in the Nordics could provide a unique complement to existing investments nationally and in the EU. The Nordic countries are highly digitalised societies with digitally literate populations, advanced industries, and developed e-governance systems, common societal traits, as well as strong university sectors and innovation capacities, providing fertile grounds for research and innovation activities with added value. For example, the Nordic countries are characterised by high levels of societal trust and it is vital that AI is developed, regulated and used in an ethical and responsible manner in order for the trust not to be undermined. In this context the Nordic region has strong potential to be a frontrunner for its own benefit as well as providing a significant contribution to European efforts in the area.
In this section the NordForsk administration outlines the objectives of a Nordic initiative and presents two scenarios for Nordic research and innovation cooperation. In the following section some recommendations for future work are outlined.
Objectives of a Nordic initiative and Nordic added value
The objective of the suggested initiative would be to fund research and innovation activities that address shared Nordic societal challenges in the context of the full-scale use and implementation of AI in the region. The initiative should contribute to mobilisation of capacity and competence building among Nordic research and innovation communities and strengthen networks and collaborations. With increased capacity and strengthened and extended networks, Nordic researchers and innovators would be more competitive in EU framework programmes.
The initiative should have significant Nordic added value to the research and innovation communities, Nordic research and innovation funders as well as the Nordic societies in general. Development of AI technology is outside the remit of the initiative.
A strategic priority of the Nordic Council of Ministers is for the Nordic collaboration to contribute to a green, competitive and socially sustainable region. Responsible and ethical use of AI will be vital in achieving these priorities as well as many of the UN sustainability goals. But to be able release the potential of technology solutions in a responsible and ethically sound way research and innovation is essential.
Capturing the many opportunities presented by AI and developing models for ethical and responsible use would be at the heart of the suggested initiative. The initiative should respond to societal challenges that the Nordic region is facing, such as ageing populations, ongoing structural developments in the labour market, polarisation and the green transition of the economy. The initiative should also address AI specific challenges with a clear Nordic added value such as the consequences of AI generated content on democratic discourse, trust in content, and digital literacy. Inclusionary perspectives should be considered with regards to systemic bias and users at special risk such as children, young people, elderly and disadvantaged people. In addition, the initiative should address various aspects associated with ethical AI, presented in various settings such as the EU High-Level Expert Group on AI Guidelines for Ethically Trustworthy Artificial Intelligence.
With these objectives in mind, the initiative would provide added value in relation to most current EU funding in the area.
Scenario 1: Open call(s) on the consequences and potentials of AI in the Nordic societies
The technological development of Artificial Intelligence is rapidly evolving, and its use and implementation is changing and developing in unpredictable patterns. AI has a very broad application potential, and it is difficult to point to specific sectors that will be more affected than others. This speaks for building in a considerable degree of openness and flexibility into the initiative. An open call would enable the creativity and knowledge of researchers and innovators to identify timely areas where new knowledge is required and would mitigate a potential risk that the initiative would miss out on essential topics or areas of knowledge that the funders could not foresee in advance. An open call could create conditions for excellence and creativity and stimulate research and innovation in a broad set of topics and application areas. An open call would ensure that the joint effort would keep its relevance throughout the process of constructing a call and reviewing applications. It would also require flexibility throughout the lifetime of the funded activities.
A considerable risk with an open call is that NordForsk would receive an unmanageable number of applications, creating an overwhelming amount of work for the administration and the reviewers as well as negatively impacting success rates. The NordForsk administration therefore proposes some actions to focus the call, ensuring relevance and value for money;
Focus on Nordic added value and the EU AI act
The involved funders in the call should develop descriptions of the Nordic added value expected from this initiative. The NordForsk administration has identified three preliminary areas of special Nordic interest based on shared values, social structures and societal challenges in the course of the work with this discussion paper. These areas are described with some more detail under the next scenario.
- The Nordic societal model
- Nordic values, languages and cultures
- Environmental sustainability
These or similar topics of special Nordic interest could also serve as keywords in a call, as in the NordForsk call on Sustainable Development of the Arctic. Applicants would need to describe the Nordic added value and the contribution to building knowledge essential for Nordic societies in the application.
In addition, an important aspect of Nordic added value would be to generate knowledge about the consequences of the implementation of the EU AI act in the Nordics. The EU AI act will come into full effect during the course of the initiative, preliminarily during the first half of 2026. There are considerable unknowns about which effects the legislation will have in general and on the Nordic countries in particular. The NordForsk administration therefore proposes that the call includes the effects of the new legislation and its implementation.
Data bias in relation to gender and equality
The availability of high-quality big data is a prerequisite for well-functioning AI. Built in bias in data is therefore reproduced by AI leading to issues of representation and discrimination. Bias should be addressed in the call when relevant.
Crosscutting strategic priorities
- Encouraging future capacity building in the area of AI through including early career researchers in key roles in proposals
- Requiring partners from four countries to create conditions for ambitious, stronger and broader multilateral research consortia with access to more data sources, wider surfaces for collaboration and networks and with significant potential for impact
- Expecting cross-disciplinary research teams
- Including stakeholders and users in the research projects
The initiative could also Include public and private sectors and public-private partnerships (PPPs).
Two stage application and review process
The first stage would consist of a short research proposal outline and the novelty of the application would be addressed in the first review stage together with an assessment of Nordic added value, the cross-disciplinary nature of the proposal, stakeholder involvement, and inclusion of early career researchers. This stage would consist of an eligibility check by the NordForsk administration and an assessment by an expert committee. A selection of the proposals would be invited to submit a full application that would be reviewed according to NordForsk regular procedure by an expert committee as well as a call committee. The call committee has the mandate to make strategic priorities to ensure a balanced portfolio.
Scenario 2: Thematic call(s) on the consequences and potential of AI in the Nordic societies
In light of the rapid development and implementation of AI technologies there are some shared areas of interest that are of critical societal significance for the Nordic countries and the Nordics as a region. It is therefore important to acquire knowledge to be able monitor and follow the developments and identify the range of consequences in these areas of particular interest. Consequences are and will be of a positive and negative nature and it will be important to have knowledge across the whole range. For example, research and innovation can be used to identify, amplify and fully capitalise on Nordic strengths and provide grounds for transfer of knowledge and lessons learned between sectors. At the same time, knowledge will be needed to mitigate risks, threats and negative consequences that are pronounced in the Nordics or threaten Nordic values and societal structures. With increased knowledge compensatory measures can be effectively developed and deployed to mitigate the effects of various risks. In short, a Nordic initiative on research and innovation can be used in the navigation of the digital transition of the economies to facilitate adaptation and stimulate resilience, using digitalisation as an advantage and human resources where they are most needed.
With a thematically focused call themes can be tailored and described with national priorities in mind, creating possibilities to pool funds from various government budgetary sources and thereby perhaps enabling more funding to be allocated to the initiative. Themes can be developed that are important for Nordic collaboration (and of particular Nordic added value) and that are complementary to national and EU efforts. With a clear anchoring in national and Nordic priorities there would be a clearer avenue to communicate results to policy and practice.
As mentioned under Scenario 1, the NordForsk administration has identified three preliminary areas of special Nordic interest based on shared values, social structures and societal challenges in the course of the work with this discussion paper. The areas of particular interest are the Nordic societal model, Nordic languages and culture, and environmental sustainability including climate change and biodiversity. As the areas to a large extent are intertwined, themes and topics are listed below in no particular order of importance.
- The Nordic region is characterised by strong democracies, high levels of trust and robust social cohesion. AI generated content is and will be used to spread misinformation and fake content that is damaging for the social fabric, undermines societal security and drives polarisation.
- Artificial intelligence has already had an impact on working life and labour markets and long-term consequences are difficult to foresee. AI will most probably continue to replace administrative tasks and generate qualified and creative content leading to job losses. However, as in previous transformations of the economy other sectors and jobs will emerge and grow and productivity can increase exponentially. Life-long learning and digital literacy will probably only grow in importance.
- The Nordic countries have high welfare ambitions and world class health care systems. AI can be used to decrease manual administration, increase effectivity, efficacy and accuracy, and tailor bespoke interventions. AI could relieve welfare systems under pressure from ageing populations, workforce shortages and increasing expectations from populations.
- The use of AI tools in teaching and learning is already widespread and it is important to generate knowledge about the effects on learning outcomes as well as on skills, creativity and critical thinking among different groups of students.
- Artificial intelligence can be deployed to support sustainability efforts, for example enhancing manufacturing efficiency, supporting the transition to sustainable energy sources and materials, and aiding the shift to sustainable agriculture. AI can also be used for monitoring purposes, surveying biodiversity and climate change. The environmental impact and sustainability of AI technologies is also an important aspect.
- Structural changes to the economy do not have an equal impact on the population, with winners and losers emerging relatively early on. A focus on inclusion and groups at risk is a central principle of the Nordic societal model and is important to counteract further polarisation.
- Large Language Models (LLM) are a prerequisite for high quality generative AI. Generative AI tools developed based on non-Nordic languages and trained on non-Nordic content might not be linguistically or culturally appropriate in a Nordic context. Similarities between Nordic languages and cultures can contribute to more relevant language models for Nordic and national conditions and there are several unique data repositories in the Nordics that potentially could be combined for this purpose.
The actions to focus the call, ensuring relevance and value for money outlined under scenario 1 are also appropriate to consider under scenario 2. Regarding the modality of the call, in terms of application and review processes, the NordForsk administration recommends that processes should be tailored to choices regarding the ambition and breadth of calls.
Conclusion
The scenarios presented here, including themes and Nordic added value, are a preliminary suggestion from the NordForsk administration. A working group consisting of representatives from interested funders could do further work identifying and developing a successful model for a joint call. An advantage with Scenario 1 could be that it would be easier to communicate as an ambitious initiative focusing on AI and it could open for perhaps more creative and novel approaches. Scenario 1 could also to a large degree build on existing Nordic strengths among the research and innovation communities, creating and strengthening capacity and competencies. Adopting Scenario 2 could on the other hand fund research and innovation in areas with common Nordic challenges with clear added value and policy relevance. Scenario 2 could potentially require some more time and resources to delineate but would also be administratively less burdensome and easier to execute.
NordForsk recommends focusing on identifying instruments and thematic scoping with the broadest possible support and ambition among the funders within the timeframe indicated in this discussion paper in the chapter below.