By Alex Leggett, Research Executive, FleishmanHillard Brussels
It appears that EU policymakers have finally woken up to the widespread impact that is certain to come from the further development of AI. Emerging technologies, such as AI, will present a challenge to policymakers as they try to catch up with the pace of development. However, this also presents an opportunity for Europe’s politicians to shape development in a policy area with little concrete legislation.
On 26 June, the European Commission’s Expert Group on AI presented its Policy and Investment Recommendations to provide a basis for the future of European AI legislation. To this point, the Commission’s 2018 strategy for ‘Artificial Intelligence Made in Europe’ represents the EU’s most comprehensive response to how it should address the opportunities and challenges posed by AI. Based on this, the European approach to AI revolves around four key areas: creating an ethical framework for AI; ensuring the availability of quality data; promoting skills and investment; and preparing the European workforce for the AI revolution.
In this article, I take a wider perspective to assess what European policymakers have thus far discussed and what steps have been taken. While much of society speculates on the risks and rewards posed by AI, policymakers will need to take concrete actions which will directly impact Europe’s approach to AI in the coming years.
Many European policymakers accept that they have “missed the boat” when it comes to being at the forefront of AI research. Indeed, it was estimated in 2016 that European external investment in AI totalled $2.4-3.2 Billion in 2016, compared to $6-$10 billion in Asia and $12-18 billion in North America. Thus, the EU has looked to ethical considerations to put its stamp on the development of AI. This focus on the ethical impact of AI has become the key strategy for the EU to place its mark in a field which has become increasingly dominated by other regions.
Therefore, the European Commission tasked a High-Level Expert Group of 52 specialists to produce a set of ethical guidelines with the aim of ensuring trustworthy AI in Europe. These guidelines, published in April, set out a framework outlining ethical principles for AI based on fundamental rights such as respect for human autonomy and the explicability of AI decision-making. Although these guidelines are non-binding, the ethical principles are likely to provide the basis for any future legislation in the field of AI and are being actively promoted by the EU in an international context.
Access to Data
One of the principal obstacles to the rapid development of AI systems is the availability of high-quality data on which machine learning algorithms can train themselves. There is a reason why it has become a cliché to say that data is the new oil which will power the economies of the digital age. The EU’s challenge, therefore, is how to ensure that AI developers in Europe are not held back by a lack of quality data sets. This is a challenge of which many European policymakers are acutely aware, as European Commissioner Margrethe Vestager noted earlier this year that the EU “need[s] the raw material” if it is to deliver its plan on AI.
Artificial intelligence is also set to be a driver for the EU’s wider ambition to increase data access and sharing across the European Union. The recently revised Directive for the Reuse of Public Sector Information will encourage EU Member States to more freely share their publicly held data, thereby increasing the availability of high-quality data to be used by AI developers. Looking forward to the future EU data agenda, we can expect that AI will provide impetus for further initiatives to incentivize data sharing in both Business-to-Business (B2B) and Business-to-Government (B2G) perspectives.
Skills and Investment
Although it is widely acknowledged that high-quality data is needed to power artificial intelligence, it is also clear that Europe’s development of AI systems will not be possible without both skilled engineers and adequate investment. A large volume of skilled tech workers is one aspect in which Europe retains a competitive advantage. Indeed, the EU, through its December 2018 Coordinated Action Plan on AI, has encouraged Member States to prioritise AI skills within their national AI strategies. However, as 240,000 Europeans were working in Silicon Valley in 2017, Europe needs to better incentivise its talent to remain in the EU.
The EU has also used its Coordinated Action Plan to reveal its strategy creating a public-private partnership which aims to increase private investment to challenge funding levels seen elsewhere in the world. Artificial intelligence has also become a priority within the EU’s next long-term budget as it commits to over €3 Billion in funding through the Horizon Europe and Digital Europe schemes. This, however, represents only a small fraction of planned investment for AI, which is primarily expected to be derived through private sector funding.
Preparing Europe’s Workforce
One thing that almost all key stakeholders admit about artificial intelligence is that it will bring great changes to European society in the form of both opportunities and challenges. Indeed, European Commissioner Věra Jourová admitted that AI “will change the way we live and work” in Europe. The challenge posed by AI to the future of work is tackled within Policy and Investment Recommendations drafted by the European Commission’s High-Level Expert Group on AI. In particular, the Expert Group recommends that the Commission establishes a European transition fund to manage the social impact of AI in the workplace with calls for extra funding to facilitate the up-skilling and re-skilling of affected workers. However, it remains to be seen if such measures will be sufficient to truly tackle the immense societal and economic impact of the AI revolution.
Europe as a Global AI Leader?
Overall, it is clear that Europe’s policymakers are now wide awake to the revolutionary risks and rewards posed by AI, and they are ready to face such challenges to make Europe a global leader. With Europe having fallen behind the United States and Asia at the forefront of AI development, it has been crucial to develop a uniquely European approach to AI. Through a focus on ethical principles, it is possible that Europe can be a pioneer in ensuring that AI will bring both a positive economic and societal impact.