Michael Laha, Václav Kopecký
Europe’s cautious approach to developing emerging technologies is hampering its global competitiveness. The European Commission’s AI strategy falls far behind China’s ambitious blueprint in several key aspects, including funding, sector-specific policies, startup incentives, and talent attraction. This article is part 2 of a mini-series to present the outcomes of the MERICS European China Talent Program 2018.
On April 25, the European Commission released its long-overdue blueprint for a European strategy to boost research in and industry applications of artificial intelligence (AI). These new measures represent a first step toward implementing an EU-wide initiative, at a time when Europe is falling behind the United States and China in AI development. While the United States is still leading in AI research and top talents, China is rapidly catching up with ambitions to overtake.
Europe cannot afford to miss out on the vast economic gains promised by AI, which has the potential to add 10 to 12 percent to its GDP by 2030. Beijing’s aggressive push in AI has been met with concern across the EU. German Chancellor Angela Merkel has called for Germany to compete with China in AI and French President Emmanuel Macron announced an AI strategy for France.
The most striking difference between Europe’s and China’s AI strategies is the level of ambition. The Chinese central government has a penchant for setting lofty goals, but its national AI benchmarks have sent local governments into a frenzy. Provincial and municipal administrations are trying to outcompete each other by pledging hundreds of billions of CNY to AI research and development. Tianjin city alone is planning an AI fund made up of public and private investment totaling 100 billion CNY (ca. 13 billion EUR) – an amount unimaginable for many European countries, let alone cities.
China’s AI industry benefits from a powerful combination of financial firepower, a favorable regulatory regime, and readily available data from nearly 800 million internet users. If Europe hopes to stay competitive, it will need to go far beyond what the EU Commission report lays out.
China scores with big vision and targeted incentives
The State Council’s Next Generation AI Development Plan of July 2017 is the cornerstone of China’s AI strategy. By 2030, China wants to become the world’s leading innovation center for AI, boasting a 1 trillion CNY (ca. 130 billion EUR) AI core industry. A three-year action plan subsequently provided concrete guidelines as to which AI technologies and sectors China could develop.
The government has implemented targeted incentives to fill China’s development gaps in AI theory and core components. A plan by the Ministry of Education addresses the problem of China’s limited talent pool by providing measures to improve AI talent training. A new government-backed research center for AI development is one of many moves to pool AI resources from academia, government and industry. Meanwhile, nearly half of all global investment in 2017 into AI start-ups ended up in China and the world’s highest-valued AI start-up, SenseTime, is Chinese.
Europe’s AI strategy puts forward fewer and less ambitious goals than China’s. By raising its AI investments to 1.5 billion EUR under Horizon 2020 – the EU’s biggest research and innovation program – the Commission hopes to stimulate a total increase of 20 billion EUR in private and public AI investment by the end of 2020. Most recently, it pledged another 2.5 billion EUR in AI investment as part of its new Digital Europe program. Yet, the Commission’s AI document admits that Europe is lagging behind other parts of the world in private investments by billions of euros.
Europe is home to a world-leading AI research community, extensive deep tech expertise, and competitive players in robotics, yet its strategy lacks guidance on how to capitalize on these strengths. Much of Europe’s high caliber talent has flocked to more lucrative markets because start-ups are often unable to scale up at home.
China’s top-down government spending does not guarantee the technological breakthroughs needed to take the global lead in AI. As with previous industrial policy, this approach will inevitably waste vast amounts of resources. Yet, it deserves special attention for mobilizing public and private players. China’s “wish list” approach to AI details the technologies the government deems essential without narrowly restricting researchers or entrepreneurs. It has rallied policy-makers, academia, and industry behind one mission and incentivized local governments to support private AI endeavors.
Europe needs a bolder approach to AI policy
If Europe wants to keep up, it must take a bolder approach to AI policy making. The Commission should devise concrete strategies to encourage AI research, applications, and investment in priority sectors such as robotics, while addressing weaknesses such as Europe’s worrying brain drain. Europe may be strong in basic research, but academic excellence and patents do not necessarily directly translate into economic benefits if commercialization of new technologies lags behind. Europe has to envision other mechanisms to attract start-ups, venture capital firms and AI talent, and bring these players together. Some of the tactics that China has rolled out could very well work in Europe. Tax incentives for companies, cash rewards for researchers, and preferential visa policies could help create a more supportive innovation ecosystem in the EU as well.
Another key advantage is that China seems to be more comfortable with the changes ahead. Policymakers and industry representatives consistently praise AI as a potential force for good. China’s population feels at ease with adopting new technologies into everyday life situations, from mobile payments to AI-enhanced e-commerce chatbots and news aggregator platforms. The European public is more skeptical: a recent survey conducted in the UK reveals a lack of trust in algorithms among civil society. In China, discussions on the socio-economic and ethical implication of AI have only just emerged, with AI researchers and policymakers calling for stronger awareness of AI safety issues.
Politicians across Europe have emphasized that the EU must not take a Chinese or American approach to AI but carve out its own path. The strategy stresses a “European way,” that, “benefits people and society as a whole.” It consists of a commitment to preparing for the inevitable socio-economic changes brought about by AI and creating an appropriate ethical and legal framework. Planned ethics guidelines will focus on the future of work, fairness, safety, security, and social inclusion.
If Europe wants to set standards it needs to lead first
The “European way” promises to be more attuned to the socio-economic consequences of AI's rise. However, the current plan is heavily imbalanced. Two of the EU Commission strategy’s three sections detail defensive measures for guarding against risks, while only one section embraces the vast opportunities that AI technologies will bring.
This will hamper the development of cutting-edge AI in the EU. Already, the General Data Protection Regulation (GDPR) has raised concerns that it could hinder AI innovation, by limiting how personal data is used and by raising the legal risks for AI companies. This means that European policymakers need to take into account that certain regulations, such as GDPR data privacy protections, may also pose additional obstacles for the development of AI technologies and the growth of companies that rely on them.
The European Union will only be able to effectively advocate for its AI standards globally if it is a leader. This will require balancing important discussions on the sustainable, safe, and just use of AI with an ambitious effort to develop top-notch AI technologies.
If European policymakers want to shape the future of AI, there are a few things they can learn from China when it comes to creating a “European way” – and selling it. Chinese government announcements on AI emanate a sense of enthusiasm about how AI will improve people’s lives and advance China’s economic growth. China may be underestimating the risks of the AI hype. But a European strategy that fails to sufficiently support European advancement in AI technology will not put the EU in a position to compete against China and determine the rules of the AI game.
About the authors
Rebecca Arcesati is a Junior Analyst at MERICS.
Caroline Meinhardt is an Analyst at MERICS.
Michael Laha is a Program Officer at the Asia Society’s Center on US-China Relations
Václav Kopecký is a Research Fellow at Association for International Affairs (AMO) and a Consultant at CEC Government Relations
The authors participated in the fourth annual MERICS European China Talent Program in April 2018, during which parts of the argumentation presented in this blogpost were developed. The authors bear sole responsibility for the content.