In a panel discussion hosted by the Center for Data Innovation, policymakers and industry representatives discussed the EU data strategy. While the panelists agreed that the EU data strategy is an opportunity to address key technical challenges in data sharing, data quality, and interoperability, many also saw opportunities to clarify how the EU will govern data infrastructure and operations.
According to Malte Beyer-Katzenberger, team leader for data policies in the unit for data policy and innovation at the European Commission’s DG Connect, to capture the potential of data, the EU data strategy aims to develop governance mechanisms that connect, accommodate, and empower the different players of a more diversified and decentralized data economy. For this purpose, common European data spaces will enable all stakeholders to participate in defining how data governance will work in practice and to lower data sharing costs—which, according to the results of the Commission’s public consultation on the data strategy, respondents agree is necessary to foster voluntary data sharing. Andreas Hartl, head of division strategy for artificial intelligence, data economy, and blockchain at the German Federal Ministry for Economic Affairs and Energy, supported the creation of sector-specific data spaces with specific governance approaches tailored for each sector’s challenges.
Panelists discussed challenges to data governance and data sharing. Hartl referred to the low availability of data, which is itself caused by the lack of incentives for stakeholders to share data, a lack of trust, and concerns over privacy and data protection. Other obstacles include the lack of interoperability, restricted access to certain data, costs of acquiring data, compliance costs, and legal security in data use. 80 percent of respondents to one of the consultation’s questions said that they have encountered difficulties in accessing and using data from other companies. Beyer-Katzenberger and Teemu Ropponen, general manager at MyData Global, mentioned how a growing number of data management companies acting as data intermediaries can help address these challenges, such as by monitoring the use of data.
Eline Chivot, senior policy analyst at the Center for Data Innovation, noted that challenges to data sharing, privacy concerns, but also issues in AI systems’ inaccuracy could be addressed by a greater focus on data quality rather than just on interoperability and data standards. Data quality is underrepresented in the data strategy, while it could create a better data market and help companies expand globally. Beyer-Katzenberger responded that data quality is challenging to implement: It is costly for businesses, and difficult for policymakers to mandate.
Companies will be important stakeholders in the implementation of the EU data strategy and as such, they should be incentivized with a voluntary framework that provides guidance, legal certainty in the markets they invest in, and a level playing field. There are many existing data sharing initiatives within industry, codes of conducts, and self-regulation frameworks. According to Jessica Lennard, senior director for global data privacy and AI initiatives at Visa, and Philip Malloch, director of economic and social policy at Facebook, there are strong incentives for companies to share data—for example to enhance their own datasets, to participate in exchanges, to seize revenue opportunities, and to develop non-profit activities based on data sharing. Companies such as Facebook and Google are involved in data transfer projects that allow users to move their data across platforms. The role of policy should be to streamline this activity and encourage the development of complementary tools, standards, and business models built through data trusts or co-labs. In addition, there does need to be alignment and synchronization between industry, policy objectives, and new regulations which might interact with existing schemes, to avoid legal complexity. Agile, flexible, co-creative regulations are a better fit to technology than overly prescriptive ex-ante regulations.
Panelists referred to the important role of individuals in the EU data strategy’s vision as active agents and empowered consumers in the data economy. With personal data spaces, and by providing more tools for real-time use of data, the EU’s framework aims to give more agency and control to individuals over their data, including to those who want to donate data for an altruistic purpose. The results of the Commission’s consultation suggest that a large share of respondents believe that law and technology should enable citizens to do so. (The Commission has now published more details.) Ropponen said that the GDPR’s data portability rights should be adjusted to make this possible. In addition, the EU data strategy aims to give individuals tools so they can exercise their rights better, but they need to understand these tools to be able to use them. Panelists concurred that education, skills, and digital literacy are critical to this purpose.
Beyer-Katzenberger and Hartl reiterated the EU’s commitment to the global free flow of data and in multilateral systems of data flows with other partners. According to Lennard, a project such as Gaia-X, if it avoids foreclosing participation from global players who might bring their expertise, resources, and ideas into the EU’s data environment, can be positive. Digital sovereignty was discussed as a way to build competitive infrastructure, but that should not lead to regimes that mandate data sharing without justification, to regulatory divergence, or to data localization which is detrimental to the free flow of data, the economy, and consumers.