In this Rise of AI Virtual Chat from 2020, Fabian Westerheide interviews Damian Boeselager, Member of the European Parliament for Volt and member of the Greens/EFA group and the budget committee, about the European Union's data strategy and its relationship to AI regulation. Boeselager recounts co-founding the pan-European party Volt in 2017 in response to Brexit, Trump's election and rising nationalist parties, and explains why Volt joined the Greens group in the European Parliament to gain committee seats, including the budget committee. The two then discuss the EU's AI white paper, led by Commissioner Margrethe Vestager, and the separate data strategy pushed by Commissioner Thierry Breton, who argued the EU needs a data strategy because algorithms cannot be trained without data. Boeselager explains the EU's proposed risk-based framework for regulating AI by how much it infringes fundamental rights, the tension between European sovereignty (data security, control of critical infrastructure) and competitiveness (relying on more advanced US cloud providers), and how the EUR 750 billion Covid recovery fund could be spent on digital and green transformation with minimum spending thresholds Boeselager is pushing for.
Key takeaways
- Boeselager co-founded the pan-European party Volt in 2017 after Brexit and Trump's election, and joined the Greens/EFA group in the European Parliament because an independent MEP has no influence.
- The EU pursued an AI strategy under Commissioner Margrethe Vestager (a risk-based framework tied to fundamental rights) separately from a data strategy under Commissioner Thierry Breton, who argued data must come first to train algorithms.
- GDPR cost Volt half its subscribers due to compliance uncertainty and legal costs, but Boeselager still calls its core data-ownership principle a genuine European success.
- Boeselager argues EU policy conflates sovereignty (data security, control of critical infrastructure) with competitiveness (needing US cloud providers, who currently lead technologically), and that these require different solutions.
- He is pushing for minimum spending thresholds, at least 25 percent each on digital, green and resilience projects, within national plans for the EU's EUR 750 billion recovery fund.
Detailed notes
Founding Volt and joining the European Parliament's Greens group
Boeselager describes co-founding Volt in 2017 with an Italian collaborator, prompted by Brexit, Trump's election, the resignation of Italy's Renzi over a constitutional referendum, and the AfD entering the German parliament, which he recalls as around 90 seats. Volt built a Ruby on Rails platform to connect new members instantly with local representatives, prioritizing a positive onboarding experience. Registering a European party required registering separately in each country, ranging from about 60 signatures in Germany to 5,000 in Portugal and 150,000 notarized signatures in Italy, which Volt could not meet. After winning a seat in the 2019 European Parliament election, Boeselager says an independent MEP is 'irrelevant' without a parliamentary group, so Volt negotiated committee seats and joined the Greens/EFA group, which let Boeselager join the constitution, migration/asylum and industry committees, and later secure a seat on the budget committee, which handles the EU's Recovery and Resilience Facility.
Separate AI and data strategies, and a risk-based regulatory framework
Boeselager explains the European Commission's AI white paper, driven by Executive Vice-President Margrethe Vestager, proposes regulating AI according to how far it infringes fundamental rights, on a five-level scale, with the strictest regulation reserved for the highest-risk uses such as autonomous weapons systems and HR selection processes. The separate EU data strategy, pushed by Internal Market Commissioner Thierry Breton, reflects his view that a data strategy must come first because algorithms cannot be trained without data. Boeselager says the AI white paper drew 168 pages of consultation feedback and that a workable framework would need graduated responses below the top tier, for example licensing requirements or mandated transparency about training data, rather than only banning the highest-risk category.
GDPR's costs for startups versus its data-ownership success
Boeselager recalls that Volt lost half its subscriber base when GDPR took effect because unclear compliance requirements and legal costs (he cites a lawyer quoting EUR 6,000 a day) can seriously harm an under-resourced organization. He argues future AI regulation must avoid similarly harming small businesses, for instance through free legal guidance, but calls GDPR's core principle, that individuals own their data, a real success that has shaped privacy standards across Europe and internationally.
Sovereignty versus competitiveness in cloud and critical infrastructure
The two discuss whether the EU data strategy amounts to building a closed 'digital fortress.' Boeselager says the white paper mixes two distinct goals, sovereignty (data security against actors like US government access under the Cloud Act, cyberattacks, and control of critical infrastructure) and competitiveness, and argues that small European companies digitizing today still need to work with Google, Amazon or Alibaba because European providers lag technologically. His preferred fix for US companies is not exclusion but requiring them to pay fair taxes locally, comply with EU privacy law, and guarantee (potentially via splitting off a legal entity) that data physically stored in the EU is not accessible to foreign governments; Boeselager says such data-residency solutions (for example hosting in Frankfurt) are already technically offered by major cloud providers. He also argues the EU should mandate interoperability between cloud providers, especially for core infrastructure-as-a-service offerings, to prevent lock-in and preserve competition.
Regional strategies and the EUR 750 billion recovery fund
Asked whether all EU member states share the same view on data, Boeselager says most countries currently lack a clear data or AI strategy, making EU-level coordination important to avoid fragmented national rules; he cites Sweden's regionalized approach, for example focusing the Groningen area on healthcare data with public funding, legal clinics staffed by GDPR lawyers, and networking support, as a model he wants replicated in Europe. On the EUR 750 billion Covid recovery package, financed through EU-level bonds and distributed mainly via the Recovery and Resilience Facility, member states submit national plans that the Commission approves; Boeselager says current rules are not strict enough and he is pushing for minimum spending thresholds of at least 25% each on digital transformation, green transformation and resilience (including healthcare) within national plans.
Q&A highlights
Asked whether requiring US firms to split into separate EU legal entities could push American companies to reward users who sign up directly with the US service instead, Boeselager says he hesitates to jump straight to mandating a corporate split; his hope is that US companies find technical and contractual ways, verifiable through monitoring that no data packages leave EU soil, to guarantee data independence from US government access without full entity separation. He adds that he is skeptical of government attempts to build a wholly separate European Amazon-, Google- or Facebook-equivalent from scratch, calling that idea a source of nervousness given his private-sector background.
Quotes
The barriers to entry and democracy are extremely high in many countries.
Damian Boeselager
How would you train your algorithm if you don't have the data to do that?
Damian Boeselager
Questions answered in this talk
Why did the EU develop a separate data strategy alongside its AI white paper?
According to Damian Boeselager in this Rise of AI Virtual Chat, Commissioner Thierry Breton pushed for a dedicated data strategy because, in his view, algorithms cannot be trained without data, making data policy foundational to AI policy.
How does the EU's proposed AI regulation classify risk?
According to Damian Boeselager, the European Commission's AI white paper proposes a five-level scale based on how much an AI application infringes fundamental rights, with the strictest rules reserved for the highest-risk category, such as autonomous weapons systems and HR selection processes.
Does the European Union want to build its own separate cloud and internet ecosystem?
According to Damian Boeselager, the EU's data strategy raises the question of a 'digital fortress,' but he is skeptical of governments trying to replicate providers like Amazon or Google from scratch, and argues small European companies still need US cloud providers to digitize competitively today.