At Rise of AI 2020 in Berlin, Damian Boeselager, a Member of the European Parliament and co-founder of the pan-European party Volt, argues that AI and data regulation should be set at the European level rather than fragmented nationally. He describes how the European Commission paired Margrethe Vestager's AI strategy, which classifies applications as high or low risk, with a data strategy from Thierry Breton aimed at enabling data sharing across industries. Boeselager lays out three theses: prioritize EU-wide rules over national fragmentation to protect the digital single market; separate digital sovereignty from competitiveness by strengthening competition law and interoperability rules rather than building state-backed European cloud competitors such as GAIA-X; and make regulation concrete and tangible, citing GDPR as a global standard that nonetheless burdened small organizations like his own. He points to Sweden's regional-cluster approach as a model for helping SMEs move from analog operations to using AI. In the Q&A he addresses GAIA-X, data-sharing mandates for big tech, and how to make industrial data usable without destroying startups' competitive advantage.
Key takeaways
- Boeselager argues AI regulation should be set at the EU level, not fragmented across national laws, to protect the digital single market.
- He pushed for a 20 percent digital target, about 135 billion euros, within the EU's roughly 750 billion euro recovery fund.
- He separates digital sovereignty from competitiveness, favoring strong competition law over building state-backed European cloud competitors like GAIA-X.
- He calls GDPR a global privacy standard but says it halved Volt's follower base due to compliance burdens on small organizations.
- He cites Sweden's regional-cluster AI strategy, linking local companies and universities, as a model for helping SMEs adopt AI.
Detailed notes
From Founding Volt to the European Parliament's AI Committees
Damian Boeselager describes co-founding Volt around three or four years earlier with an Italian and a French colleague, aiming to build the first genuinely European party, before winning one of Germany's seats in the European Parliament. He now sits on the parliament's newly founded special committee on AI, the industrial policy (ITRE) committee, and the budget committee. He recounts how the new European Commission moved quickly on AI: Margrethe Vestager proposed an AI strategy early on, and Thierry Breton, brought in as the French commissioner after the Parliament rejected Emmanuel Macron's original candidate, paired it with a data strategy. Boeselager says this pairing is smart, because regulating AI without a strategy for sharing data leaves companies without data to train algorithms on. Vestager's approach sorts AI applications into high risk, such as autonomous weapons systems, and low risk, such as a Spotify recommendation list, with stricter rules reserved for high-risk cases.
Thesis One: European Regulation, Not National Fragmentation
His first thesis, which he says departs from an earlier speaker, Dorothee Bar, is that AI regulation should be set at the European rather than the national level, to avoid further fragmenting the EU's digital single market and weakening already-strained European competitiveness. As an example of pushing this agenda, he says he fought to include a '20 percent digital target' in the roughly 750 billion euro European recovery fund, meaning about 135 billion euros of the instrument he works on should go toward digital transformation.
Thesis Two: Separating Sovereignty from Competitiveness
His second thesis is to clearly separate digital sovereignty and strategic autonomy from competitiveness. He is skeptical of initiatives to build a federated European cloud infrastructure to compete with the three big US cloud providers, referencing GAIA-X, which he says originated from the German industry ministry. Rather than steering companies toward European providers, he argues the EU should apply strong market regulation and competition law to any provider operating on European soil, requiring interoperability, portable APIs and safeguards so data cannot be transferred to other countries, for example under the US Cloud Act, regardless of whether a company chooses AWS, Google Cloud or Azure.
Thesis Three: Concrete, Tangible Regulation, Learning From GDPR
His third thesis is that regulation should be concrete and tangible rather than overly complex. He calls GDPR a genuine success in setting a global standard for data privacy, but says it halved Volt's follower base because smaller organizations struggled with recertification and audit requirements, and argues future rules need built-in support to help small businesses and startups understand and apply them. He points to Sweden's AI strategy, built around regional clusters linking local companies, universities and utilities around concrete use cases, as a model for a bottom-up approach, including funding for local 'legal clinics' to resolve legal questions as companies move from analog operations toward using AI.
Q&A Highlights: GAIA-X, Forced Data Sharing and Industrial Data
Asked to explain GAIA-X, Boeselager describes it as an attempt to federate European cloud providers so they can jointly compete for tenders, particularly public procurement, though he says he would rather set clear pan-European rules for any cloud provider than build a public-sector competitor. Asked how to get big tech to share data with smaller players, he favors competition-law remedies, citing an active US case over Amazon's market access, over a general legal obligation to publish data. An investor in the audience asks whether mandatory data sharing would undermine startups whose product is built on proprietary annotated data; Boeselager says he has not found a workable way to force sharing except where withheld data blocks market entry, and instead favors pooling scarce AI talent and data-architecture expertise across similar companies. On industrial data, which he says is held in a highly decentralized way across SMEs, he argues the next transformation depends on shifting company mindsets from hoarding data to sharing it, and on industry data standards that make sharing fast once companies agree. Asked for his wish from the European Parliament, he repeats his call to prioritize European over national AI regulation to preserve competitiveness against China and the United States.
Quotes
GDPR really managed to set a global standard when it comes to data privacy and data ownership.
Damian Boeselager
The GDPR came into place and really halved our whole followership.
Damian Boeselager
If you don't have the right regulation or strategy of how to share data, you won't have any data to train your algorithms on.
Damian Boeselager
Questions answered in this talk
Should AI be regulated at the national or European level?
According to Damian Boeselager at Rise of AI 2020, AI regulation should be set at the European level rather than through national laws, because national fragmentation would further complicate the EU's digital single market and weaken European competitiveness against the US and China.
What is GAIA-X?
According to Damian Boeselager at Rise of AI 2020, GAIA-X is an initiative, originating from the German industry ministry, to federate European cloud service providers so they can jointly compete for tenders, especially public procurement, against the large US cloud providers.
Should big tech companies be forced to share their data with smaller competitors?
According to Damian Boeselager at Rise of AI 2020, he favors using competition law to address cases where withheld data blocks market entry, but has not found a workable general legal mandate to force companies to share data, especially where it would undermine startups built on proprietary annotated data.