Using AI is necessary for solving key societal challenges, staying competitive, and in general preserving and developing a democratic society with a high quality of life. The objective of An AI Strategy for Sweden is therefore to develop Sweden’s capabilities to strongly benefit from the potential of AI.
Like many other countries, Sweden faces urgent grand challenges such as climate change, increasing energy demand, an aging population, and threats to our democracy and security. The consequences of the demographics alone turn the usage of AI and modern technology into a necessity, as the welfare sector otherwise would need a significant share of the newly graduated workforce at the same time as there are limits to tax raises. Today’s solutions are simply not the solutions that will build the future of society.
In the private sector, many companies are seeing rapidly increasing international competition and signs of a coming dramatic disruption, often driven by new possibilities enabled by AI. To preserve and strengthen competitiveness, Sweden should adopt policies that spur its private sector and, not least, large exporting corporations to accelerate their investments in AI.
Countries that early on have invested heavily in AI research and talent through research-focused strategies (e.g. Canada) now face a crucial choice to invest beyond research and experimentation to enable adoption or lag behind. While others have to change their strategies, Sweden on the other hand has an opportunity to acknowledge that the benefits from AI come when the technology is put to use in companies and organizations, and implement an adoption-focused strategy from the beginning.
Sweden should act now and apply a bold strategy aligned with democratic values, incentivizing substantial investment by companies and organizations and aiming at generating and spreading the benefits across sectors and society.
An AI Strategy for Sweden, therefore, encourages Sweden to recognize the necessity of AI and apply a fully adoption-centric strategy that ensures significant and broad value across sectors, contributes to a prosperous democratic society, strengthens our national security, and enables Sweden to have a positive global impact while navigating potential risks.
The heart of such a strategy should be bold leadership focusing on value, cross-sector collaboration, and collective investments, fueled by international relations and strong frontrunners in both the private and public sectors boldly leading the way.
An AI Strategy for Sweden has been written to guide politicians, business leaders, decision-makers, civil servants, or changemakers with ambition and responsibility to lead and develop Sweden, the Swedish ecosystem, or individual organizations and companies towards a positive future through one of the most transformative times in history.
This first version of An AI Strategy for Sweden is intended to catalyze actions by leaders in national and regional governments, government agencies, corporations, and public organizations.
It can not be emphasized enough that Sweden’s success in adopting AI requires broad and collective actions from a wide range of companies and organizations. It is not down to the government alone, although the government plays an important role as a key enabler.
This Strategy has been developed by AI Sweden, the national center for applied AI with the mission to Accelerate the use of AI for the benefit of our society, our competitiveness, and everyone living in Sweden, with feedback and input from selected experts in Sweden, Canada, USA and Singapore. The Strategy is also inspired by the strategies of EU, Canada, the USA, the UK, Germany, and Singapore.
An AI Strategy for Sweden has been written by the leadership team and other team members at AI Sweden led by Martin Svensson, Managing Director, and Mikael Ljungblom, Director of Public Policy and International Relations.
While AI Sweden is the sole responsible organization for the strategy, the following contributors have given input and feedback in their individual capacity:
Executive Director, Pan-Canadian AI Strategy at CIFAR
President at Dakota State University and one of the commissioners in the National Security Commission on Artificial Intelligence
Director, AI Innovation at AI Singapore
Vice-Chancellor (Rector) at Luleå University of Technology
Chief Strategy Officer, SAAB
Director General of Vinnova, the Swedish Innovation Agency
Chief Strategy Officer, board of Halland hospital group
Secretary General, the Swedish Confederation of Professional Employees
VP and Head of Data Office, R&D at AstraZeneca and Chairperson in AI Sweden
Many others have contributed with inspiration and perspectives.
A future Swedish society making the most of the potential of AI should draw on existing strengths centered around democracy, social stability, a highly developed welfare state, high levels of equality, and a business environment that fosters innovation and entrepreneurship.
Based on who we are as a country, our values, and our strengths, the future Sweden seizes the opportunity of AI to solve key societal and business challenges for the benefit of a prosperous society characterized by democratic values and high quality of life. This is powered by a robust, highly attractive, and ambitious Swedish AI ecosystem that has a positive global impact on people and the planet.
The Vision can also be expressed as Sweden striving to become a worldwide benchmark for how AI is embedded nationally and generates sustainable value throughout businesses and society with a positive impact globally.
As an overall indicator of immediate progress, Sweden should aim to be ranked among the top 10 countries in the Global AI Index (current ranking is 17) by 2025, at the latest.
A broader framework for benchmarking and measurement based on the Global AI Index and this Strategy is suggested further down.
This Strategy is based on a strong sense of urgency and an analysis summarized in a set of fundamental standpoints and perspectives that must be recognized by any politician, business leader, decision-maker, civil servant, or changemaker with ambition and responsibility to lead and develop Sweden, the Swedish ecosystem, or individual organizations and companies.
These standpoints and perspectives also underpin the heart of the strategy and the Key principles below.
The last years have clearly shown that the use of AI at scale and profound breakthroughs happen in the private sector, not least in the USA. Companies with access to relevant data, resources, leadership, and talent constantly increase their lead through a positive reinforcement loop of attracting more talent, generating more value, increasing investments, attracting more talent, and so forth.
The trend is clear, organizations and companies that understand AI the best, also increase their investments the most. As a consequence, the gap between the best and laggards continues to widen. As the value and potency of AI typically scale exponentially, the importance of not falling behind cannot be emphasized enough.
To stay relevant, Sweden as a nation, and Swedish companies and organizations, should engage and learn from the best on all levels.
Reference: HAI AI Index Report 2023
The speed at which AI is being developed is unprecedented, recently demonstrated by generative AI (on language, images, and video). New services have reached hundreds of millions of users and started to change knowledge work in an exceptionally short period of time.
This pace of change can be expected to continue and force companies and organizations to adapt. The pace of development in the private sector, where there is access to real-world data and significant resources, also challenges academic research on AI methodologies and technology to stay relevant.
To stay adaptive and strengthen Sweden’s competitiveness, the Swedish AI ecosystem needs to lift to a higher level. This means that new ways of working, new incentives, new funding processes, and new ways of collaborating should be identified and deployed.
As noted in the introduction, even countries that have invested in AI research and talent through research-focused strategies early on(e.g. Canada) now face a crucial choice to invest beyond research and experimentation to enable adoption or lag behind.
Mature and powerful AI technologies are already today freely available or accessible as ready-to-be-used services and tools. Although generative AI is used to substantially improve the quality and productivity across organizations and processes, the potential value of available technology is far from fully realized.
While AI technology is typically openly available, data held by companies and organizations is often proprietary and constitutes untapped value and opportunity for strengthened competitiveness and societal solutions. Hence, a strategy aiming at accelerating the use of AI broadly should prioritize investments driving adoption and value-creation based on such proprietary data.
In academia and research institutes, there is also an untapped potential for adopting existing AI technologies and accelerating and enhancing research in various fields (e.g. medicine, biology, materials, etc).
As the adoption of AI happens broadly, research related to the consequences on society becomes increasingly important. Such research could include research on the economic potential, societal value, consequences on the labor market, and the safety of AI systems.
Determined and knowledgeable leadership understanding the potential value and risks of AI, setting direction, and making concrete investments is critical for any nation, organization, or company that wants to broadly benefit from the opportunities of AI. Lack of leadership leads to missed opportunities and increased risks of AI being used in ways not aligned with responsible AI principles, not beneficial for society, or misaligned with other objectives.
Further, the competence of an organization will determine how well and sustainably AI solutions will be built and deployed. Leaders understanding the potential of AI therefore immediately and forcefully strive to increase knowledge level and competence broadly among its leaders and workforce in a purposeful way on a strategic, tactical, and operational level.
Lastly, investing in capabilities and competence is essential but not enough. Without a shift in mindset, those investments will not render expected returns, what took us here is not what will lead us into the future. Leaders need to have an ambidextrous approach, both working strategically and visionary to take the nation and organization on a long-term journey while simultaneously creating value and solving problems using currently available AI solutions. This needs to be communicated and embodied by executive leaders and decision-makers.
The demand for AI skills & talent is growing, and a highly skilled workforce is a critical long-term need of any AI ecosystem. Frontrunner nations, regions, and companies have for many years made, and continue to make, significant efforts to attract, develop and train talent to build and use AI.
The issue of skills & talent in an ecosystem is enormous, and includes education of future workforce on many levels, training of existing workforce (re-/upskilling), attracting and retaining talent to the field of AI, and attracting international talent to Sweden. The responsibilities for the various aspects are highly distributed which needs to be considered when setting a strategy for how to address the needs.
It’s equally important to not only focus on technical skills and talent to drive AI adoption and initiatives. AI is a cross-disciplinary field requiring skills as well in change management, interaction design, legal, business models, communication, innovation management, and many more context-dependent skills.
Unlike many technologies, the deployment and adoption of AI technology and solutions come with value creation and synergies across all sectors of society.
Moreover, very few organizations/companies, or even nations, can build enough capability and competence themselves. Extensive collaboration is therefore essential and Sweden should recognize the need to invest in the building of AI ecosystems by facilitating and incentivizing collaboration.
For Sweden especially, with the high decentralization of the public sector (e.g. 21 regions, 290 municipalities, 350+ government agencies), collaboration becomes even more important as very few organizations individually are capable of building and managing necessary infrastructure, data, and solutions. To really scale the benefits of AI across both organizations and the country, significant activities and investments compensation for the decentralization must be put in place.
“The research community, industry, and Government play important, interconnected roles to enable the development and deployment of AI solutions. It is critical for Singapore to enhance partnerships across these stakeholders to strengthen our collective capabilities and drive our overall national AI effort.”
“With new opportunities, risks, and threats to prosperity and security at stake, the promise and peril associated with this foundational technology are too vast for any single actor to manage alone. As a result, cooperation is inherently needed to equally mitigate international security risks, as well as to capitalize on the technology’s potential to transform enterprise functions, mission support, and operations.”
As set out in the Executive summary, An AI Strategy for Sweden encourages Sweden to recognize the necessity of AI and apply a fully adoption-centric strategy that ensures significant and broad value across sectors, contributes to a prosperous democratic society, strengthens our national security, and enables Sweden to have a positive global impact while navigating potential risks.
The heart of this adoption-centric strategy should be bold leadership focusing on value-creation, cross-sector collaboration, and collective investments, fueled by international relations and strong frontrunners in both the private and public sectors boldly leading the way.
To guide politicians, business leaders, decision-makers, civil servants, and changemakers, the strategy includes a mindset that should permeate, guide, and align actions, investments, and initiatives associated with this strategy.
The mindset is set out in the form of the following Key Principles.
Relevant actions, investments, and objectives should be based on and mapped against these principles for Sweden to quickly see positive effects and over time maximize the long-term impact throughout the ecosystem.
The Key Principles should apply generally but might need adjustment to fit a specific context or leadership role.
Ambitious frontrunners, i.e. the early adopters and leading companies and organizations investing and demonstrating real value by AI, already exist in both private and public sectors.
Such frontrunners should be recognized, enabled, and incentivized to move even faster and generate more value given that they share generously with others in the ecosystem. This would result in Sweden quickly reinforcing a pay-it-forward culture encouraging responsibility, investments, talent attraction, and collaboration, as well as contributing to making Sweden a global example of how to create substantial value through AI adoption.
This strategy's focus on frontrunners is based on the trend that breakthroughs typically appear in companies and sectors where AI is applied to real problems, often in service and product development.
Given the pace of development internationally and that many of the objectives are best managed on an international level, Sweden as a nation as well as organizations and companies should actively seek partnerships with like-minded countries, leading companies, academia and research institutes, and non-profit organizations, where there is a ground for mutual learning and development.
Sweden and Swedish companies should especially build relations with leading global companies and ensure access to knowledge, infrastructure, foundational models, services, and products critical for the development and deployment of AI solutions.
International collaboration is often challenging and generating actual, mutual, and sustainable value takes time. Sweden’s investments in international collaborations should therefore preferably be pragmatic, focused and strive towards a high density of interactions with selected ecosystems.
Moreover, it should be mentioned that international collaboration must be developed considering the current geopolitical, trade, security, and competitive situation. This can be difficult to manage on an organizational level, and government guidance and facilitation should be provided.
Leaders throughout society should invest in upskilling their workforce and attracting new talent to their respective organizations, whether that is within government, private companies, public organizations, ministries, government agencies, or academia.
For Sweden as a nation, dedicated efforts are necessary on many fronts, not least
From a governmental perspective, this represents a significant challenge. Clear direction and coordination of the entire educational system should be prioritized as well as investments to scale initiatives quickly increasing the overall competence in the ecosystem.
In parallel, Sweden must make it easier for international talent to discover, arrive, study, work, live, and remain in Sweden, as well as incentivize and enable the private and public sectors to recruit talent internationally.
The development of strategic and important knowledge, technology, and solutions for sovereignty, independence, and scaled adoption should be incentivized and built-in broad collaborations, and made available as widely as possible. Strategic knowledge, technology, and solutions (and other relevant resources) must evolve and should include national computational infrastructure, foundational models, privacy-preserving methods, access to data, data-sharing mechanisms and structures, and AI adoption tools for leaders.
AI Sweden (among others) has demonstrated that the willingness among international and Swedish companies and organizations to jointly invest in knowledge, technology, and solutions enabling the adoption of AI is significant. When incentivized by government funding and structured on a truly collaborative basis with strong sharing mechanisms, the return on investment can be significant.
Furthermore, to fully benefit from AI, strategic investments addressing specific sectors’ underlying prerequisites are necessary. This specifically applies to the public sector, where investments in digitalization, data platforms, and data governance are still required.
By building on the Swedish values centered around democracy, inclusion, transparency, and diversity, Sweden has the opportunity to scale and govern the use of AI in a responsible way. AI systems should be designed, operated, and governed in a manner that respects laws, human rights, values, and cultural diversity. This includes mitigating risks such as bias and discrimination, as well as risks related to security and safety.
As a principle, responsible use of AI should be approached and ensured by active experimentation and exploration while developing AI solutions in practice, and by fully recognizing the need for diversity and inclusion of stakeholders, domain experts, end users, and AI developers in all parts of the process.
Moreover, leaders in any organization should ensure that employees have access to, and proficiency in, available modern tools independent of for example their age, native language, disabilities, educational background, or where they live in Sweden. Related to AI, this extends to providing prerequisites for everyone to use AI responsibly and minimize the risks that groups of people are left behind.
Responsible use of AI also means that we have a responsibility to use AI where appropriate and to address our time’s most pressing global challenges.
To maximize the benefits of AI, Sweden should scale initiatives that catalyze collaboration across all sectors and regions. This should include a strong ambition to compensate for the high decentralization of the public sector through investments in collaboration and sharing between similar public organizations.
Facilitation of collaboration should include incentives for consortiums to build joint advanced solutions, structures, and ways to enable access to real-world data and in general build on mechanisms that allow fast pace and adaptivity.
Sweden should also consider investing in one or two “moonshots”, e.g. large-scale initiatives where many actors come together to solve a grand challenge.
An AI strategy for Sweden points to Sweden’s opportunity to use AI to solve key societal and business challenges for the benefit of a prosperous society characterized by democratic values and high quality of life.
As stated in the Vision, Sweden should strive towards becoming a worldwide benchmark for how AI is embedded nationally and generates sustainable value throughout businesses and society with a positive impact globally. What would constitute such a worldwide benchmark that can inspire others to follow?
Ultimately the benchmark must demonstrate that real value is generated broadly throughout society. Measuring only enabling factors, the size of investments, or the progress in some sectors would not be enough.
Existing frameworks (for example the Global AI index ) are mainly focused on enablers for AI such as talent, infrastructure, and innovation (research and development) and to some degree investments. These could serve as indicators of use and adoption, but will not provide a holistic view.
To follow up on Sweden’s progress, a framework combining existing well-established indicators and additional indicators to further measure the actual use and value creation is therefore relevant. There are few established metrics for use and value creation and such indicators need to be further developed.
More work is required to develop a pragmatic and detailed framework of indicators, but the following high-level structure of three main pillars is suggested
Enablers - the status of necessary prerequisites for broad adoption
Investments - the level of investments of various types
Use and value creation - quantification of solutions deployed and the estimated value
Fundamental prerequisites for broad adoption in a country, sector, or organization include operating environment, access to talent and competence, infrastructure and other relevant technology, and solutions.
Indicators related to investments are very relevant as they both give a perspective on the belief in the future value to be realized, but also serve as a proxy for the current leadership commitment regardless of level. Metrics should be on government investments as well as investments on a sector level.
Given the importance of collaboration, the framework of indicators should also measure investments in cross-sector collaboration, and international collaboration as well as investments in knowledge, technology, data, and solutions related to sovereignty, independence, and scaled adoption.
An adoption-focused strategy needs indicators of actual use and value creation. Further development is needed to measure the ecosystem at large concerning use and value creation specifically by frontrunners per sector and in total. As value and impact created by AI can be direct or indirect, proxy indicators such as the number of AI models or solutions deployed might be relevant.
Going forward, a framework and its indicators need to be further defined, specifically indicators on Use and value creation. Preferably indicators should be developed to match already publicly available data. A first step would be to measure the current state of the Swedish AI ecosystem with regard to the developed framework and set up tracking mechanisms to follow the progress.
As an overall and immediate objective, Sweden should aim to be ranked among the top 10 countries by 2025 at the latest.
An AI Strategy for Sweden is AI Sweden's suggested common ground for continued discussion and action. So what does the journey to a defined and adopted national strategy with clear priorities, investments, and allocated responsibilities look like?
As it addresses a vast number of leaders throughout business and society, this journey is both highly important and challenging.
Therefore, let’s agree that it is our actual ability to take action that determines the progress, not the choice of words. It is our firm belief that bold actions guided by this strategy will propel us towards the suggested Vision.
Passivity is coupled with great risks in terms of diminished competitiveness, less tax revenue, highly qualified jobs being created elsewhere, an overburdened public sector, and an overall significantly less prosperous society.
On the other hand, if we succeed in taking action throughout society, we are looking at vast opportunities in attracting talent and foreign investments, increasing tax revenues, creating highly skilled jobs, providing excellent service and welfare for our inhabitants, and creating a key foundation for a prosperous future.
We think that a few key questions will further aid us in making priorities, taking concrete action, and distributing responsibilities.
Again, it's worth emphasizing that Sweden’s success requires bold leadership, broad collaboration, and significant investments. Individual organizations or companies might prevail on their own, but for a prosperous Sweden, no entity, or the government for that matter can carry the responsibility alone.
The above questions are formulated for the national level, the government being a key enabler, and similar questions will be equally important for a hospital, municipality, or company management to answer and act upon.
Launching this first version of An AI Strategy for Sweden, AI Sweden is committed to taking responsibility and contributing to the journey to a defined and adopted national strategy with clear priorities, investments, and allocated responsibilities, learn more here!
For questions about the strategy as such, reach out to Martin Svensson or Mikael Ljungblom.