Reforms to boost AI in the EU: an overview of TBI’s ideas
As EU leaders debate digital sovereignty in Berlin, a new Tony Blair Institute paper lays out how Europe can stay competitive in the AI age – by reforming regulation and capital markets, boosting compute and energy, and exporting its digital government stack.
As the European leaders are currently discussing the concept of European digital sovereignty in Berlin, the Tony Blair Institute for Global Change (TBI) released a new paper “Europe in the age of AI: How Technology Leadership Can Boost Competitiveness and Security”, which outlines necessary reforms to reshape the EU’s regulatory framework, ensure the right technical capabilities are in place, accelerate AI adoption, and find better ways to export the European digital government stack.
At first glance, the paper provides concrete, structural recommendations for the EU, without falling into the simplistic narrative about public procurement quotas that, in our view, unfortunately, often dominates the debate on the EU’s digital sovereignty.
The problems
The paper highlights many structural problems in the EU – at the governance, regulatory, and implementation levels – that hinder AI adoption, development, and the export of the “European digital stack” outside of the EU.
The authors say that fragmented markets and complex regulations hinder businesses from scaling, while the European innovation ecosystem is held back by “chronic underinvestment”, topped off with “some of the world’s most expensive energy prices”.
The report is critical of a simplistic perception of digital sovereignty:
“Today, European debates too often conflate digital sovereignty with autarky: the belief that we must build domestic alternatives for every technology. This is economically infeasible and strategically confused: true sovereignty is not about ownership, but about leverage and choice.”
Solution #1: reforming digital regulations, building better access to capital, and reforming labour markets
The first area that the paper suggests reforming is not surprising at all: the TBI calls for “reforming Europe’s regulations and decision making to create an innovation-friendly digital market and ensure that Europe-wide questions have Europe-wide responses”, as the “existing digital regulatory framework makes it costly for businesses to innovate, scale and compete globally, and the AI Act illustrates these challenges.”
The paper further argues that the EU should utilise the upcoming Digital Omnibus to “simplify reporting, harmonise enforcement and streamline regulation across member states”, and fast-track implementation of a Savings and Investment Union and a 28th regime.
Aside from that, the paper calls for the creation of a continent-wide stock exchange and for the modernisation of labor markets – unfortunately, the latter hasn't received the attention it needs in the EU.
Solution #2: physical and digital foundations, compute capacity, and affordable energy
The paper references the EU’s current initiatives, such as the EU Gigafactories and EuroHPC JU, and does not go into detail on whether current initiatives and the distribution of AI factories will reach the necessary scale. It also, very welcomely, calls for creating conditions to “attract large-scale private investment in AI infrastructure”.
Energy-wise, the paper advocates for the acceleration of a European energy union and the “launch of a continental energy program” to coordinate the development of new nuclear power plants and expedite permitting for renewables, grids, and storage.
Solution #3: realistic incentives to accelerate AI adoption
In terms of AI adoption, the paper focuses on three areas: the availability of data; a regulatory framework fit for accelerated AI adoption; and reforming how European academia works.
First, the paper suggests accelerating AI adoption by building a trusted, interoperable ecosystem based on open-source innovation. The TBI wants to see the upcoming Data Union Strategy “unlock strategic data sets” and use public procurement and compute-credit incentives to facilitate market pull.
Side note: mentioning the reform of the General Data Protection Regulation (GDPR) is, at this point, politically dangerous in Europe, and the TBI hasn’t mentioned it explicitly. However, if the upcoming Data Union Strategy is complemented by a bold Digital Omnibus reform encompassing the GDPR for AI, we could see the change the EU needs so much.
Second, the TBI suggests creating sector-specific AI adoption regulation packages that provide regulatory clarity and fast-track approval paths.
Third, the paper calls to “align university systems with the demands of the modern innovation economy to attract researchers and strengthen Europe’s talent base”. The authors argue that European universities are constrained by “rigid governance, outdated incentives and uncompetitive salaries”, which in turn drives top researchers and AI talent abroad. To change that, the TBI suggests that governments, funding agencies, and universities should reform career structures, increase autonomy, and offer globally competitive pay, as well as flexible pathways between academia and industry.
This is a welcome suggestion, but given what we know about some of the problems in EU universities, we believe autonomy and additional funding should not be a given for everyone, at least not for those who are underperforming. Those that are already delivering clear, tangible results, however, should be granted both greater autonomy and the necessary funding.
Solution #4: Look for better ways to export the “European digital stack”
The fourth area in the paper is “strengthening Europe’s global tech engagement to project influence and values worldwide”.
Concrete suggestions include launching a “coordinated strategy to export Europe’s digital government stack to the world” through initiatives such as Global Gateway. The paper further focuses on establishing European regional tech hubs across the world, staffed with technologists and diplomats, to promote a European tech stack. Third, the paper suggests establishing a “European Investment Acceleration Mechanism” which would support investors in navigating regulatory requirements for strategic projects.
Side note: reforming the Global Gateway has been on the European Commission’s agenda, but some critics argue that it would be easier to start over rather than reform it in a way that would actually make EU companies competitive with US or Chinese companies in continents such as Africa or Latin America. Yet, realistically, no alternative mechanisms exist, and the EU should certainly invest in attracting technologists to support the global export of its digital government stack. Currently, the process is often fragmented, as companies are left to manage their exports independently, while many diplomats lack the necessary expertise to facilitate it fully.
An acceleration mechanism to navigate complex rules is also a welcome addition. Although some would (traditionally) argue that this would create an additional layer of complexity, similar mechanisms are already applied to foreign investment projects across some of the EU countries, greatly facilitating the practical launch of projects. The same principle can and should be applied to strategic AI-related projects.
