In the past few years, the European Union has adopted a suite of digital regulations aimed at shaping the behaviors of a wide range of businesses, such as global technology companies, financial service providers, and small businesses that interact with their customers online. Some of the latest regulations to enter into force in 2024-26 include the Digital Markets Act, an ex ante digital competition policy framework; the Digital Operational Resilience Act (DORA), impacting financial service providers and their information and communications technology (ICT) vendors; and the Artificial Intelligence (AI) Act, which affects firms developing and deploying AI systems.
Several leading Asian economies are today contemplating or advancing similar EU-style digital regulations – so it is time to ask, have the EU’s digital regulations served Europeans? Will the new regulations promote what Asian countries are all going for – an innovative and competitive economy with vibrant startups and small and medium-sized enterprises (SMEs), digital transformation, and productivity?
This paper seeks to answer these questions through a review of data on the performance between Europe and the U.S., as a benchmark, in producing startups and digitizing SMEs, and assessing the potential impacts of the DMA, DORA, and AI Act on Europe’s economy and startup ecosystem. The paper catalogs a significant divergence between the EU and the U.S. in terms of innovation and startup formation, one that will likely be exacerbated by the new digital regulations.
The paper calls for Asian policymakers to pursue alternative regulatory frameworks that can address legitimate policy concerns in the digital economy, while also enabling innovation, investment, and inclusive growth.
The main findings are as follows:
There is a significant divergence between Europe and the U.S. in startup formation and venture capital, R&D, and firms’ technology adoption, all of which undermine Europe’s economic growth. The U.S. has eight times more unicorns, or companies with a valuation over $1 billion, than Europe, and of the world’s 50 leading technology companies, only four are European, and none of these were created in the past 50 years. European unicorns also gallop away: between 2008 and 2021, nearly 30 percent of Europe’s unicorns relocated their headquarters abroad, mostly to the U.S.. As the AI era unfolds, the U.S. outpaces Europe about four to one in producing AI startups and eight to one in providing VC funding in AI startups. Europe also lags behind the U.S. in digitizing businesses that are in traditional industries (manufacturing, mining, professional services, and so on). In addition, U.S. firms across size categories also invest more in R&D, a key driver of productivity growth, than European firms.
A considerable body of academic literature traces Europe’s lackluster performance in technology adoption and commercialization over the past three decades to its labor, business, and digital regulations that cost companies, lower the payoffs from technology investments, and limit companies’ ability to optimize their labor force. As Europe’s digital regulations come into effect, they risk exacerbating these problems while doing little to promote technology adoption and startup activity in Europe.
European Union leaders themselves are expressing concerns about the negative effects of the continent’s burgeoning regulatory stock on innovation and startups. In January 2025, the European Commission published the Competitiveness Compass that outlines steps to drastically reduce the regulatory and administrative burdens, especially for SMEs, and in May 2025, the European Commission considered postponing the implementation of the AI Act due to industry concerns about its lack of clarity.
The various new digital regulations, as they are rolled out, entail immediate compliance costs and potential fines for firms targeted by them, amounting, conservatively, to 0.2-0.5 percent of European GDP. Per-firm compliance costs with the DORA, discovered in surveys with financial services firms, are by now €500,000-1 million for firms with more than 500 employees. McKinsey estimates that the largest financial entities would spend €5 million-€15 million each in DORA compliance.
Compliance costs are disproportionately higher for SMEs and startups. These costs can hit startups hardest: a recent interview-based study with 23 European AI startups found that these firms anticipate annual compliance costs with the standards-related activities of 10-20 percent of the management’s time and some €100,000 per year in dedicated compliance personnel, a significant expense for a startup with, say €1 million capital raised.
There can also be various second- and third-order effects from the DMA, DORA, and AI Act that have far-reaching implications for European businesses, such as:
​
-
Lowered incentives for firms to invest in technologies and R&D, both because of expectations of lower returns from technology and because of diminished resources left to invest. For example, in an indicative survey of 200 European financial services firms for this report, the surveyed businesses expect to reduce their non-DORA-related technology expenses on average by 1-5 percent and their R&D by 6-10 percent to deal with the costs of DORA compliance. A larger share of small firms report steep spending cuts than do large firms – which could suggest that DORA sets back especially smaller firms’ innovation.
-
New barriers to entry to startups and SMEs – which face higher fixed costs of implementing, for example, a quality management system and standards to deal with the AI Act. Smaller financial services firms that may rely more heavily on third-party IT vendors than large financial firms face a new task of verifying that these vendors are in compliance with DORA.
-
Higher prices for the targeted companies’ customers. For example, the majority of the 200 financial services firms surveyed plan to raise the prices they charge their customers due to the costs of dealing with DORA, typically by 3-5 percent. Costs may also be indirect and unexpected. For example, because of the the DMA a Google Search for hotels in Europe now elevates third-party aggregators in search results instead of highlighting direct links to hotels, European hotels have experienced 30-40 percent drop in bookings coming through their own sites as opposed to bookings via aggregators – which implies a transfer of revenue in the form of commissions from hotels to the aggregators.
-
Lower quality and variety of services available to the customers and users of digital services. For example, in a new survey with 5,000 European consumers, 61 percent report having to search up to 50 percent longer for relevant online content than before the DMA, and as many as 59 percent would pay to regain direct access to targeted firms’ own services, including rich shopping results that enable product and price comparisons. The DMA is also impacting advertisers that may face new challenges in building specific audiences or running personalized advertising campaigns. In the survey, some 39 percent of Europeans report less personalized ads after the DMA came into effect. A new study finds that these various search frictions and lowered personalization are costing European SMEs across sectors, with annual revenue losses per worker up to €1,122.
Asian policymakers should be mindful of these effects that risk weighing on Europe’s innovation and competitiveness, and that now also concern Europe’s own policymakers. Rather than rushing to adopt EU-style laws simply because the EU has pursued them, governments should first identify whether there is a problem in their economy that needs to be addressed, and if so, whether it should be addressed through regulation instead of using other alternative approaches.
For example, in the area of digital competition policy, Asian policymakers should thoroughly assess whether their domestic markets actually face the same competitive challenges that prompted the EU's actions that resulted in the DMA. Asia’s digital markets are still evolving, with local players competing effectively against global tech companies, suggesting that heavy-handed regulation might stifle innovation and economic growth rather than promoting it.
In financial services, there are already meaningful templates to follow: the Financial Stability Board’s toolkit for financial institutions and the Basel Committee on Banking Supervision’s consultative document are the key ones. These address third-party risk management, for example, through risk assessments, due diligence, contracting, monitoring, and termination. In addition, many financial hubs, such as New York, have published risk management rules and principles that are more targeted and risk-based than DORA.
In the area of AI, EU-style regulations should also be seen as just one plausible model. For example, Japan’s May 2025 AI law, which calls for collaboration between the government, businesses, and researchers to promote the development of AI in Japan, is a very useful alternative model to the prescriptive EU AI Act. Another model is presented by Singapore, which has pursued a balanced AI regulation, promoting companies’ adoption of AI standards and expanding compute and data for use in R&D. The UK similarly has promoted pro-AI innovation regulation, including through its ambitious 2025 AI Opportunities Action Plan and the pragmatic and pro-innovation approach to the use of data to promote AI – such as promoting sandboxes and consensual audits of high-risk businesses on how they use personal data for AI.
​
​​
​​
​