AI Sweden, the national center for applied AI, was inaugurated in February 2019 with the firm ambition to accelerate the use of AI for the benefit of our society, our competitiveness, and for everyone living in Sweden. Four years later, AI Sweden consists of over 100 partners across all sectors, involves thousands of individuals, operates from 8 cities, and demonstrates leadership and collaborations across areas of strategic importance to all organizations that have the ambition to generate sustainable value using AI.
By collaborating on key enablers for AI adoption, our partners have leveraged their accumulated knowledge to further accelerate the use of AI and solve key business and societal challenges. AI Sweden serves as the platform for directing efforts and investments where they can have the most impact. This collaborative approach has started to yield significant results, and this report showcases examples of our progress.
2022 marked significant global progress in AI, not least in the development of generative AI models. By using existing AI technologies, we are already able to address critical societal challenges, create innovative products, increase productivity, and promote a resilient, democratic society. While the awareness of AI as a radically changing force in society has never been larger, few organizations can realize its full potential on their own. Partnerships, collaboration, and shared resources are becoming increasingly critical.
We look forward to realizing the future impact of the knowledge and resources that we are building collaboratively and truly scale. Make sure to get involved – your experience, skills, and ideas are what makes AI Sweden.
Daniel Gillblad, Co-Director Scientific Vision
Martin Svensson, Co-Director Operations
Johanna Bergman, Head of Strategic Initiatives
Peder Blomgren, Chairman of the board and Vice President and Head of Data Office, R&D AstraZeneca
AI Sweden in collaboration with RISE and WASP WARA-Media and Language trained GPT-SW3, the first large-scale, generative language model developed specifically for Swedish and the Nordic languages. An early version of the model was shown at the Nobel Prize museum – the first time the general public had the chance to interact with a large, generative Swedish model. The model is currently being evaluated in concrete use cases in both the public and private sector.Read more
AI Sweden is an inspiration in building Large Language Models. AI Sweden is ahead of us in Norway, and your openness and willingness to share has been very helpful for NorwAI in our own work to do the same.
Rolf Dyrnes Svendsen, Communications Manager at NorwAI
Sahlgrenska University Hospital, Region Halland, and AI Sweden are exploring how to ensure legal robustness in innovation projects together with IMY, the Swedish Authority for Privacy Protection. This entails completely new ways of working to enable mutual learning while giving the supervisory authority IMY tools and knowledge around innovation processes.Read more
For Sweden to be able to lead the way for sustainable digitalization, innovation must be combined with strong privacy protection. When working to increase the knowledge of privacy and data protection issues among innovation actors, we see that there is a great demand for in-depth guidance in specific innovation processes. This means that we as a supervisory authority need to develop our working methods.
Karin Lönnheden, Chief of Staff at IMY
My AI is the first personalized platform designed specifically for the AI community and industry. A 'Linkedin for AI', if you will. The community-driven platform has been developed by AI Sweden and was launched in April 2022. To date, more than 2 500 individuals - representing over 250 organizations - have set up an account and shared close to 800 resources and 165 use cases.To My AI
For us, collaboration is now necessary. We are 290 municipalities with the same operational processes – together we can develop efficient operational support. My AI enables this collaboration; a platform for sharing experiences, inspiration and system effects.
Frédéric Rambaud, digitaliseringsstrateg at Strängnäs kommun
AI Sweden and the European Space Agency (ESA) collaborate to set up Φ-lab @Sweden that will focus on edge learning in space and supporting innovation initiatives, research calls and educational programs, ultimately leading to rapid knowledge transfer across industry sectors. Two visiting ESA researchers are currently placed at AI Sweden’s offices in Stockholm.Read more
The secondment to AI Sweden has allowed us to learn about the Swedish space sector and meet many of the actors in it. These exchanges and collaborations have broadened our research perspective and we are now building on this to jointly push the frontier of AI onboard spacecraft.
Pablo Gómez & Gabriele Meoni, researchers at European Space Agency
Leadership and ambition are absolutely essential for successfully adopting AI. Through the AI Sweden partnership leaders are sharing best practices, insights, and tools.
Leaders at all levels are engaged through AI Sweden’s networks, site visits, and knowledge exchange opportunities. For example, 16 corporate partners are participating in the Executive AI Accelerator to identify and discuss joint challenges. Moreover, it is clear that both Swedish and international politicians and decision-makers are curious about our approach to accelerating AI. During the year, we welcomed the Swedish Minister for Digitalization at the time, Khashayar Farmanbar, Erna Solberg, leader of Høyre in Norway, and Kaja Kallas, Prime Minister of Estonia, as well as other interested parties.
The annual Future of Democracy summit and events bring together leaders and decision-makers from different sectors of society to jointly address the question of how to modernize Sweden using new technologies while safeguarding democratic values.Read more
Over the past few years, we have learned what it takes to apply AI and are connecting thought leaders in strategically important areas throughout the partner network to build initiatives and operationalize AI. The focus is on making sure that AI methods can be used at scale in critical systems and real-world scenarios, and that AI systems can be developed efficiently and in a principled manner. By leading the way in this development, we ensure that Swedish organizations are building a capacity for what is to come.Read more
AI Sweden is working together with leading AI ecosystems globally to build avenues for knowledge transfer and concrete collaboration. Through AI Sweden’s presence at Mila in Montréal, Canada, partners gain access to world-leading research leading to several research and startup exchanges. Moreover, AI Sweden’s European presence has been greatly strengthened as we have initiated regular gatherings for executives at applied AI organizations in Europe. In addition, the number of global partners in our network has increased, resulting in access to competence around the world.
Very few organizations are able to fully benefit from AI if working on their own. Collaboratively building knowledge, tools, resources, labs, and models unlocks substantial value. The following examples showcase the strength and impact of collaboration within the AI Sweden network.
Region Halland, Sahlgrenska University Hospital, Karolinska University Hospital, Region Örebro Län, Region Västerbotten, Örebro University, and Halmstad University are working together on critical challenges to accelerate the implementation of information-driven healthcare and AI.
AI Sweden’s partners develop shareable knowledge and resources that cluster around five critical enablers for applying AI – Data and infrastructure, Organization and culture, Technology, Ecosystem, and Expertise.
This first enabler pinpoints access to high-quality data and infrastructure for training, evaluating, and deploying AI models.
Data Factory and Edge Learning Lab allow partners to contribute to and explore infrastructure solutions and train AI models. After exploring in a shared lab environment, organizations are better prepared to make strategic decisions about what solutions to invest in next.Read more
Many sectors face similar legal challenges related to AI and organizations therefore benefit greatly from collaborating. For example, 14 of our partners collaborate to identify legal needs with regard to data in innovation projects.
Also, Region Halland, Sahlgrenska University Hospital, and the Swedish Authority for Privacy Protection have explored a new way of collaborating in a regulatory pilot around federated learning.
Not all organizations have the data, infrastructure, and competence necessary for solving their challenges. By sharing data within the partner network of AI Sweden, innovation and research are accelerated. During 2022 two new road datasets were donated by RISE and Zenseact.All datasets
Access to data is not enough, data quality and data readiness are equally important. Several public sector organizations work on a joint project increasing data readiness. Hackathons have been one successful way of accelerating the use of data, during the year partners and AI Sweden organized hackathons on both road data and space data.All projects
Do you want to engage?
For organizations to adopt AI, it is necessary to change ways of working, processes, and structures – addressing the whole organization and its culture is therefore the second critical enabler. AI Sweden's partners are therefore investing in tools for change management and building collective resources.
Throughout the year, 12 organizations conducted an AI maturity assessment to benchmark their AI readiness and set their respective roadmaps for the work ahead. The assessment itself generates (sometimes surprising) insights and actionable outcomes. Together with Region Skåne and Region Västra Götaland, we are now exploring ways of systematically implementing AI based on the results from the assessment.Read more
I now understand that it’s not about the actual [AI maturity assessment] survey. The value lies in the discussion around the statements. We should and we will use the maturity ladder as reference for our work with AI.
Therese Furhoff, Stabschef at Östgötatrafiken
During 2022, Centiro, Region Gävleborg, Sahlgrenska University Hospital, the Swedish Tax Agency, Västra Götalandsregionen, Zenseact, and Östgötatrafiken co-developed an AI Vision white paper to support leadership teams that want to set an AI vision for their organization. This is one example of how knowledge from multiple organizations can be packaged and spread in the partner network to accelerate the overall adoption of AI.Read more
Do you want your organization to take the next steps on the AI journey?
Reach out to Conny Svensson.
The third enabler for applying AI is focused on the technology. Therefore, AI Sweden in close collaboration with partners has set up strategic initiatives around technologies that will be key for operationalizing AI. And almost 90% of the partners believe that at least one of the strategic areas is relevant to them according to the annual partner survey.
AI Sweden together with RISE and WASP WARA-Media and Language is developing GPT-SWE, the first large-scale language models for Swedish (pre-released in 2023) and a key resource for Sweden and especially the Swedish public sector. This work has put the natural language understanding group at AI Sweden and, in extension, Sweden on the international map of initiatives that develop large generative language models. During 2023 hundreds of individuals are evaluating the models on real use cases.Read more
Want to get involved?
Open lab environment with high-impact use cases exploring different aspects of training AI models in a decentralized way. Decentralized AI and federated learning approaches are quickly becoming one of the most important areas of applied AI, mainly due to efficiency and privacy reasons. The initiative has quickly gained national and international interest.Read more
Want to know more?
Every organization using AI needs their own ecosystem of collaborators. Through networks, collaboration platforms, and co-working opportunities the process of building this fourth enabler is accelerated within the AI Sweden partnership.
Events, workshops, seminars and networks enable matchmaking and knowledge sharing between peers. AI Sweden partners are gathering in both sector-specific and cross-sector networks. Another example is My AI, a community-driven platform for knowledge exchange, inspiration, and matchmaking. My AI also has an extensive use case library with 170 AI use cases – and growing.
To learn more
The fifth enabler is about expertise. Here, collaboration is key for ensuring that Sweden and each organization have access to talent and the right competence for using AI. AI Sweden partners are co-creating talent programs to boost and attract talent as well as courses for upskilling.
Want to get involved?
Does your organization want to level up its use of AI? Make sure to get involved, use our shared resources, and contribute to building new ones – your experience, skills and ideas are what makes AI Sweden