While the United States and China continue to dominate the field of Artificial Intelligence (AI), the European Union (EU) has a globally revealed comparative advantage in several sophisticated AI-related sectors, according to the latest AI Watch Index.
According to the report, EU investment in AI increased by 39 percent between 2018 and 2019. If this rate of growth continues, the EU will have surpassed its target of investing €20 billion in AI each year for the next decade.
Despite this, the EU continues to lag behind the United States and China in terms of global AI. The United States has the most AI economic players (13,770 organizations). It is followed by China (11,362 players), with the EU coming in third with 5,933 players.
The United States has a comparative advantage in several AI areas, including AI Services, Audio & Natural Language Processing, Autonomous Robotics, and Connected & Automated Vehicles.
China’s AI landscape is particularly strong in terms of new patent filings. However, the authors point out that recent Chinese policies have resulted in lower quality standards, which has led to an expansion in patents. Another factor giving China AI clout is its access to big data, which is the fuel for AI development.
The EU is a leader in AI Services and robotics, with the latter category including not only industrial trade but also the development of Autonomous robots. These must be able to operate in a highly complex environment, interacting with other machines or humans. Infrastructure, software, and platform services are all examples of AI Services.
The EU is also relatively strong in frontier research, ranking second in the number of AI publications in international conferences only to the United States, which is twice as active in this area as the EU. Projects funded by the European Commission establish the EU as a key player in the network of collaborations focused on AI research, allowing a wide range of EU economic players to participate in the AI landscape.
Another area in which the EU Member States are very active on the international stage is standardization. For example, 18 of the 33 members of the International Organization for Standardization (ISO) and the International Electrotechnical Commission (IEC) Subcommittee on AI are European national standardization bodies, with 13 actively participating. Standardization is critical for ensuring interoperability as well as managing AI risks related to safety and fundamental rights.
The Index is the result of the JRC’s AI Watch which tracks the European Union’s industrial, technological, and research capacity in AI, Member States’ AI-related policy initiatives, AI uptake and technical developments, and AI impact.
The AI Watch Index analyses multiple indicators related to the development of AI across five dimensions: a global perspective on the AI landscape, industry, research and development (R&D), technology, and societal aspects. The geographical focus is on the EU Member States, but it also includes comparisons with major global AI powerhouses such as the United States and China, among others.
The AI Watch Index findings are now available on the JRC’s newly redesigned AI Watch, which serves as a portal to AI-related research. All quantitative indicators, organized by dimension, that describe the EU’s performance and positioning in the field of artificial intelligence are available in the form of interactive charts, factsheets, and datasets.
AI-related EU actions
The European Commission launched many AI-related policy initiatives.
The 2018 Coordinated Plan on Artificial Intelligence called for increased cooperation among Member States, Norway, Switzerland, and the Commission in areas such as increasing investment, making more data available, developing talent, and ensuring trust.
The European approach to artificial intelligence is divided into two components: excellence in AI and trustworthiness in AI. Its purpose is to ensure that AI development adheres to rules that protect people’s safety, fundamental rights, and the functioning of markets.
Finally, the EU has proposed an AI Act that would implement harmonized rules. It categorizes specific AI applications into four levels: unacceptable risk, high risk, limited risk, and minimal risk.
