Asset by Undraw — adapted by Author

2020 and the European National Strategies for Artificial Intelligence

A Summary of Existing AI Strategies and Draft Strategies on AI Entering 2020

This article is a result of examinations of a variety of different national AI strategies thus attempting to get an overview of Europe in the beginning of 2020 in January. There was an idea to be helpful in giving a few pointers to help understand the variety of strategic initiatives that have surfaced. You will see that I am lacking a few central strategies, however I will update these in the coming weeks ahead.

International Collaboration on AI in Europe

Firstly, the EU has been one of the driving forces in encouraging their member countries in defining their national AI strategies.

The National AI Strategies in Europe

Belgian AI Strategy

The recent report AI 4 Belgium was published in 2019 and the authors want it to be a first version of an overarching Belgian Artificial Intelligence Strategy. The aim is to Put artificial intelligence and its implications at the top of the Belgian (federal and regional) political agenda. They aim to inspire the public debate on the required actions to undertake. Encourage the development and deployment of human-centred AI. Within it they suggest that Belgium should set up a new learning deal; develop a responsible data strategy; support private sector AI adoption; they propose to position Belgium as Europe’s AI lab; and improve public service and boost the ecosystem. They suggest this approach will require an investment of at least EUR 1 billion by 2030.

Estonian AI Strategy

The perhaps most colourful and vivid strategy is from Estonia with its Kratt report. At the same time Estonia is recognised as one of the countries in Europe that has been best at bringing technology into government (particularly with blockchain). Their strategy has a clear outline not only of big words, yet it additionally has a specific set of projects listed to help understand how Estonia is going to proceed.

Finland’s AI Strategy

Mika Lintilä the Minister of Economic Affairs in Finland appointed a steering group in May 2017 to figure out how they could become one of the world’s top countries within the field of Applied AI. In October 2017, Finland was the first European Union country to put a national action plan on AI into writing. This seems quite a lot earlier than most other countries in Scandinavia. At around that time they were scheduled to release their final report April 2019. However they also released a report at the time called Finland’s Age of Artificial Intelligence that touched upon their strengths and weaknesses in AI with eight specific recommendations to turn the country into a global leader. This led to the establishment of the Finish Centre for AI, an AI accelerator and of AI in the public service. A second report focused on work and AI. A wonderful initiative that has resulted from Finland’s work on AI was mentioned in one of my first articles defining AI. It was also partly what sparked further interest for me into the field of AI: namely elements of AI. This reached 1% of the Finnish population, and is now being adopted into all EU member state languages. The final report on Finland’s AI Strategy is called Turning Finland Into a Leading Country In The Age of Artificial Intelligence and was released in March 2019. This calls for a public discussion on the ethics of AI, continuous capacity building and bold choices followed-up by investments. Ethics seems to be a strong focus in this document. It presents different cases and has interviewed several international experts. Lastly it states that Finland is the most educated nation and can therefore be seen to have a competitive advantage in these ‘winds of change’. There seems to be a new Artificial Intelligence Steering Group that will be working towards 2025. Another initiative for piloting is AuroraAI, a national programme for implementation of artificial intelligence.

German AI strategy

In the 2019 federal budget allocating a total of €500 million. Up to and including 2025, Germany intends to provide around €3 billion for the implementation of the Strategy. It wants to empower resaerch centres and establish new ones; create 100 professorships for AI; drive forward a Franco-German research network; create funding opportunities for venture capital; establish a German AI observatory; fund in-company based test beds for AI applications; and establish a Digital Work and Society Future Fund to get the message out and to promote multidisciplinary social technology design. Strengthening research through innovation competitions and European innovation clusters. Enabling integration into business (also for SMEs) by fostering the founding of new AI businesses and supporting the changing labour market. Strengthening vocational training and attracting skilled labour and using AI in public administration. Making data available and facilitating use; adapting regulatory framework; and setting standards. Alongside this national and international networking while engaging in dialogue with society and develop framework for policy action.

Italian AI Strategy

In September 2017 the Italian AI Task Force launched the Agency for Digital Italy (AGID). The task force includes 30 direct members and around 450 community members from many sectors. One initiative was Observatory on Artificial Intelligence, which aims to analyse AI-related public conversations on social networks through technology. In March 2018, the agency released a White Paper called “AI at the service of citizens,” edited by the AI Task Force. This came to life after a consultation with hundred public and private subjects In July 2019 a draft of the National Strategy for Artificial Intelligence was released. The Italian Ministry of the Economic Development issued the national strategy. The strategy stresses public and private investments in AI and related technologies; strengthening the ecosystem of research and innovation in the field of AI; supporting the usage of digital technologies based on AI; strengthening the educational offering at every level, to bring AI to the service of workers; exploiting the potentials of the data economy; consolidating the legal and ethical framework that regulates the development of the AI. It also argues the importance of promoting awareness and trust in AI systems among citizens; boosting the public sector and make public policies more efficient; and fostering the European and international cooperation for responsible and inclusive AI.

Lithuanian AI Strategy

The Lithuanian Artificial Intelligence Strategy: A Vision of the Future was published in April 2019.The strategy in Lithuania places ethical and legal concerns highly with specific actions to address these. The two main hubs are in the cities Vilnius and Kaunas. The AI strategy in Lithuania focuses on their core areas, one of these is manufacturing as well as the integration of AI across all economic sectors. They recommend to increase the usage within private, public and targeted areas of their economy. In the education they want more focus on STEM (Science, technology, engineering, and mathematics) and supplemental coursework that teaches AI programs that did not traditionally require it. Most of the money for AI in Lithuania comes from public money according to the report. They want to finance a centre for AI, and act strategically to manage data better particularly in the public sector.

Polish AI Strategy

Poland signed the EU declaration in October 2017. The key goal of the artificial intelligence development policy in Poland for the years 2019–2027 is to provide Poland with an important place in the global economy. They have several initiatives related to data and took the lead in the Visegrad Group (V4 – Czech Republic, Poland, Slovakia and Hungary) to establish a common position on AI. The policy (Artificial Intelligence Development Policy in Poland for 2019–2027) forms part of the planned new Polish productivity strategy. It is not fully a national strategy, however is angled as a pathway towards this particular effort. It was developed by the inter-ministerial analytical and editorial team of the Ministry of Digitization and the Ministry of Entrepreneurship and Technology. Activities focus on (1) ecosystem framework, (2) organisation and management, (3) data, knowledge and competences, financing and infrastructure.

Spanish AI Strategy

Spain does not yet have a national AI strategy, however what it does have is a strategy written by Ministry of Science, Innovation and Universities, 2019 and the Artificial Intelligence Task Force (GTIA, Grupo de Trabajo de Inteligencia Artificial). This strategy is called Spanish RDI Strategy in Artificial Intelligence and will be the basis for the future National Artificial Intelligence Strategy, which will allow for the coordination and alignment of national investments and policies. It lists six priorities: (1) organisational structure that allows for impact measurement; (2) establish strategic areas for RDI; (3) define actions that allow for transfer of knowledge; (4) develop a system to promote training across disciplines and vocations; (5) develop digital ecosystem and infrastructures; (6) analyse the ethics of AI from an RDI perspective. Their recommendations is to launch the national AI strategy, use AI in public administration and education — to have an impact with AI in the labour market to attract or recover talent. They cite Roland Berger “ Joining the dots — A map of Europe’s AI Ecosystem”, the four most important AI countries in Europe are the United Kingdom, France, Germany and Spain, which contribute 60% of start — ups, laboratories and communities 33 of the 30 countries analysed (EU, plus Switzerland and Norway).

Swedish AI Strategy

In May 2018 Sweden released its strategy a National Approach for Artificial Intelligence. It acts as a guiding document for all actors in Sweden to align towards the realisation of AI benefits to welfare and competitiveness. To do this the report argues that Sweden needs to train more people in the field of AI as well as increasing basic and applied research in addition to ensuring sustainable AI (ethical, safe, reliable and transparent). Since this the government has started some new initiatives such as training and an AI science park. You can read their national strategy from 2018 here. An English executive summary of the Vinnova report can be found here. The government recently invested 40 million Swedish Kroner ($4,2 million) in training within AI.

The Netherlands AI Strategy

There was a strategic action plan for artificial intelligence launched late 2019 from the Dutch government. The Strategic Action Plan for Artificial Intelligence (AI) was launched on 8th of October by the Secretary of State Mona Keijzer (Ministry of Economic Affairs) on behalf of the Cabinet at the Humanity Hub in The Hague. With more than a hundred concrete actions such as intensive cooperation with industry and science and by investing public funds, the government will use the social and economic opportunities of artificial intelligence (AI) for the Netherlands. This is stated in the Strategic Action Plan for Artificial Intelligence that was sent by State Secretary Mona Keijzer (Economic Affairs and Climate) to the House on behalf of the government. The action plan is aimed at exploiting the opportunities of AI and includes financing for researchers and entrepreneurs. Unlike many other national AI strategies, the action plan also focuses on essential framework conditions for the successful application of AI such as: (1) training; (2) data use; (3) digital infrastructure and; (4) the protection of citizens’ rights with appropriate ethical and legal frameworks. The cabinet is already investing tens of millions of euros in AI every year. In 2019 it will be 64 million euros. The ambition is to double this public part in order to develop this technology into concrete applications together with companies and knowledge institutions. Currently, 65 parties, including the Ministry of Economic Affairs, are working together on this in the Dutch AI Coalition . This public-private partnership was launched today and forms the basis for AI’s Dutch approach with its strategy. The coalition aims for a multi-year public-private investment of a total of two billion euros in seven years. Their plan has three tracks track 1: Exploiting social and economic opportunities; track 2: Creating the right conditions; and track 3: Strengthening the foundations.

United Kingdom AI Strategy

The United Kingdom has no clear ‘national strategy’ and in a 183-page report, “AI in the UK: ready, willing and able?” it is mentioned that perhaps it should. However the closest and often referenced is the policy paper Sector Deal for AI last updated the 21st of May 2019 and released on March the 6th 2018 by the UK government. Sector Deal for AI was led by Business Secretary Greg Clark, thus its tone and focus is largely on productivity and business-oriented research. They set a few grand challenges to put UK at the forefront of industries of the future (1) AI and data revolution; (2) world leader in mobility; (3) a shift to clean growth; (4) meeting the needs of an ageing society. Their key policy moves include: raise total R&D investment to 2.4% of GDP by 2027; increase the rate of R&D tax credit to 12%; invest £725 million in new Industrial Strategy Challenge Fund programmes to capture the value of innovation. However it closely outlines the expansion of large companies (Google and Amazon) and investment (such as venture capital) into the UK.

What Countries Have AI Strategies?

If you did not see all countries in this list that is certainly the case, because not all countries have national strategies.

So who has national strategies or drafts so far and what countries do not?

I created two lists of the countries that have AI strategies and those that do not.

  • Belgium: yes.
  • Bulgaria: no.
  • Croatia: no.
  • Republic of Cyprus: no.
  • Czech Republic: yes.
  • Denmark: yes.
  • Estonia: yes.
  • Finland: yes.
  • France: yes.
  • Germany: yes.
  • Greece: no.
  • Hungary: no.
  • Ireland: no (launch early 2020).
  • Italy: yes (draft released).
  • Latvia: no.
  • Lithuania: yes.
  • Luxembourg: yes.
  • Malta: yes.
  • Netherlands: yes.
  • Norway (not an EU member, but participating): No (to be released the 14th of January 2020)
  • Poland: yes (as far as I can tell).
  • Portugal: yes.
  • Romania: no.
  • Slovakia: no.
  • Slovenia: no.
  • Spain: yes, on its way.
  • Sweden: yes.
  • Switzerland (not an EU member, but participating): no.
  1. Belarus: no.
  2. Bosnia and Herzegovina: no.
  3. Kosovo: no.
  4. Macedonia: no.
  5. Moldova: no.
  6. Russia: yes (10th of October 2019).
  7. Serbia: no (but soon).
  8. Ukraine: no.

Sidenote: Global Changes in AI Policy

Based on the AI Policy page at Future of Life Institute we can see some changes from the outlined landscape between the 10th of June and the 12th of January.

Retrieved the 10th of June 2019 from Future of Life Institute
Retrieved the 12th of January 2020 from Future of Life Institute

AI Indexes

The 2018 AI Index report from Stanford University.

AI Policy and Ethics at Student at University of Copenhagen MSc in Social Data Science. All views are my own.