‘… Europe is a unique aspiration. […] It is an aspiration of a world full of new technologies and age-old values’, Ursula von der Leyen, then incoming President of the European Commission, wrote in her political guidelines in 2019. Since then questions of new technologies and European values have been at the forefront of political discussions in Brussels and member states regarding Artificial Intelligence (AI), including preparations for the forthcoming AI Act and recently adopted Digital Services Act. These discussions have not only addressed technocratic questions of economic indicators and legal instruments but also involved soul-searching and reflections on European identity: What is Europe? What does Europe stand for? And how does Europe want to project its identity and power to the rest of the world?
Policy frames and discourses surrounding AI are at the core of an ongoing research programme which I have been leading over the past few years and which had led to a number of publications on AI governance, politics and policy (Ulnicane et al 2021a, 2021b, 2022; Ulnicane 2022a, 2022b).
Europe’s power: Single Market based on values?
In my recent publication on AI in the European Union (Ulnicane 2022a), I examine an emerging EU policy on AI, drawing on Europe as a power debate in the European Studies literature. I analyse recent EU documents on AI, asking how the EU is projecting its power in the field of AI vis-à-vis other countries and regions, especially the United States and China. Is it a Normative Power Europe relying on attractiveness of its values, as defined in the EU Charter of Fundamental Rights and Article 2 of the Treaty on European Union, namely, human dignity, freedom, democracy, equality, the rule of law and respect for human rights, including the rights of persons belonging to minorities? Or do we rather see elements of a Market Power Europe exerting its global influence by regulating access to its Single Market?
Close reading of AI policy documents launched by the EU institutions since 2017 reveals elements of both – Normative as well as Market Power. In line with an idea of Normative Power Europe, a lot of emphasis is put on developing human-centric AI that serves human needs and aligns with European values and fundamental rights. A major step in this direction has been an adoption of Ethics Guidelines for Trustworthy AI in 2019. At the same time, EU policy intentions also include elements of Market Power Europe calling for binding regulation, which are followed up by ongoing preparations for the forthcoming AI Act. Moreover, elements of Normative Power and Market Power in AI policy are interconnected, where binding regulation is presented as a way to enforce European values.
However, the EU’s activities in this field have also received criticism. For example, Ethics guidelines for Trustworthy AI have been portrayed as a potential ethics washing where talk about ethics serves as a way to delay binding regulation (Ulnicane et al 2021b). Moreover, critics have pointed out strong presence of business interests in the EU AI initiatives and questioned if EU funding for new technologies always lives up to its own values. Furthermore, some of the key EU policy documents on AI omit addressing issues of defence AI, which the EU itself is funding. While the implementation of EU political intentions to support human-centric AI remains an important issue to be followed-up in the years to come, recent cases of the General Data Protection Regulation and Digital Services Act can serve as examples how the EU’s focus on protection of its values is advanced through binding legislation.
Europe & global AI race: racing for what?
Emerging EU policy for AI draws not only on well-know values but also on some well-known political discourses such as transformation, revolution, and global competition (Ulnicane et al 2022). In particular, the discourse of global competitiveness is well-known in European integration. Already since the 1950s and 1960s concerns about the technological gap and Europe lagging behind the US have been a powerful driving force for the emergence and expansion of EU research policy. While this discourse of global competitiveness has received important criticisms over the years about depicting technological development globally as a zero-sum game where one country/ region wins and others lose, its persistence is remarkable. A new version of it can be seen in AI policy. EU policy documents for AI mention how AI development takes place ‘amid fierce global competition’ and how the investments in AI in the EU are lagging behind the US and China.
While admitting a strong global competition in AI among the US, China and Europe, the EU tries to position itself as a Normative Power, emphasising its values-based, human-centric and ethical approach to AI. In addition to global competitiveness and ‘new space race’ discourse in AI (Ulnicane 2022), there is also interest in global cooperation initiatives. The EU policy demonstrates a strong interest in global cooperation initiatives, in particular when they can help to promote the EU’s trustworthy and human-centric approach to AI. The EU has contributed to a number of international initiatives, such as the development of the OECD ethical principles for trustworthy AI, the Global Partnership on Artificial Intelligence, and EU-US Trade and Technology Council.
The main ideas from this research have been discussed in a number of invited talks. You can watch a recording of my recent talk ‘Can Artificial Intelligence be governed? Multiple challenges and some opportunities’ on YouTube.
EU policy for AI remains a moving target. Ongoing policy and technological developments are co-shaping each other with new initiatives and capabilities to be expected in the short and long term. More research is under preparation to examine these ongoing developments. Some of it will be discussed at the panel ‘Power, Politics, and Policy of Artificial Intelligence’ at the ECPR conference this summer. We look forward to discussing new insights on this topic then and on other occasions.
Dr. Inga Ulnicane is Senior Research Fellow at De Montfort University, UK. Her research focusses on science, technology, and innovation governance, politics and policy. She has published on topics such as AI governance and policy, dual use, European integration in research and innovation, Grand societal challenges, and international research collaboration.
Ulnicane, I., W. Knight, T. Leach, B. C. Stahl and W.G. Wanjiku (2021a) Framing governance for a contested emerging technology: insights from AI policy, Policy and Society 40(2): 158-177 https://doi.org/10.1080/14494035.2020.1855800
Ulnicane, I., D. O. Eke, W. Knight, G. Ogoh and B. C. Stahl (2021b) Good governance as a response to discontents? Déjà vu, or lessons for AI from other emerging technologies. Interdisciplinary Science Reviews 46 (1-2): 71-93 https://doi.org/10.1080/03080188.2020.1840220
Ulnicane, I. (2022a) ‘Artificial Intelligence in the European Union: policy, ethics and regulation’, in T. Hoerber, I. Cabras and G. Weber (Eds) Routledge Handbook of European Integrations, Routledge, pp.254-269. https://doi.org/10.4324/9780429262081-19
Ulnicane, I. (2022b) Against the new space race: global AI competition and cooperation for people. AI & Society https://doi.org/10.1007/s00146-022-01423-0
Ulnicane, I., W.Knight, T.Leach, B.C.Stahl and W.G. Wanjiku (2022) ‘Governance of Artificial Intelligence: Emerging international trends and policy frames’, in M.Tinnirello (Ed.) The Global Politics of Artificial Intelligence. CRC Press, pp.29-55. https://doi.org/10.1201/9780429446726-2