EU Digital Deregulation: Efficiency vs Distribution in Europe’s Regulatory Push (Bruegel 2025)

Why the EU Digital Deregulation Push Demands Scrutiny

The European Commission has embarked on one of the most consequential regulatory shifts in the EU’s digital history. Through a series of Omnibus packages and a landmark Digital Omnibus proposal released in November 2025, Brussels is systematically easing the compliance requirements of its flagship digital laws — the GDPR, the Digital Markets Act, the Digital Services Act, the Data Act, and the AI Act. The stated objective is to boost European competitiveness in a global technology race where the continent is falling dangerously behind.

But a penetrating 2025 Bruegel policy brief by Mario Mariniello argues that this deregulatory push suffers from three critical weaknesses: it lacks robust evidence, it obscures distributional trade-offs, and it adopts a dangerously narrow objective of replicating the American model. The brief provides an analytical framework that every policymaker, executive, and citizen affected by European digital regulation should understand. For organisations navigating this rapidly shifting regulatory landscape, understanding the full depth of the analysis through interactive policy research can accelerate strategic decision-making.

The Competitiveness Gap Driving EU Digital Deregulation

The numbers driving Brussels’ urgency are stark. In 2024, only 4 of the world’s top 50 technology companies were European. European AI startups raised a combined $12.8 billion in funding compared to $80.8 billion for their US counterparts — a ratio of more than six to one. On Stanford’s Global AI Vibrancy Index, France — the best-performing EU member state — scored just 22.5 out of 100, compared to the United States’ commanding 70.

These figures have fuelled a narrative that European regulation is strangling innovation. The argument, popularised by the Draghi competitiveness report and embraced by the current Commission, holds that the EU’s precautionary regulatory approach — establishing rules before markets mature — has created a compliance-heavy environment that disadvantages European firms against lighter-touch American and Chinese competitors.

There is genuine substance to this concern. GDPR compliance costs are estimated at up to €500,000 for small and medium enterprises and up to €10 million for large organisations. These fixed costs disproportionately burden smaller firms, potentially increasing market concentration as larger companies absorb compliance expenses more easily. Research by Peukert et al. (2022) found evidence that web technology service markets became more concentrated after GDPR’s implementation — precisely the opposite of what European regulators intended.

Efficiency vs Distribution: A Framework for Digital Regulation

The Bruegel brief’s central contribution is an analytical framework that separates the efficiency effects of digital regulation from its distributional effects — a distinction the Commission’s deregulatory proposals consistently fail to make. Mariniello introduces a model where regulation functions as a single production factor that generates two outputs: Total Factor Productivity (TFP), representing competitiveness and economic value creation, and Digital Fairness (DF), encompassing privacy, safety, trust, and security.

Efficiency improvements occur when regulation increases the combined value of TFP and DF — for example, when completing the Digital Single Market reduces fragmentation costs while maintaining consumer protections. Distributional effects, by contrast, occur when changes to regulation shift value between these dimensions — increasing competitiveness at the expense of privacy, or enhancing safety at the cost of innovation speed.

This framework reveals a crucial insight: much of what the Commission presents as efficiency-enhancing deregulation is actually distributive in nature. When the Digital Omnibus proposes broadening the definition of “non-personal data” to allow more extensive AI training, it is not simply removing an inefficiency — it is transferring value from individuals’ privacy rights to firms’ data-processing capabilities. When the AI Act enforcement timeline is pushed from August 2026 to December 2027, the efficiency gain for firms comes at the cost of delayed safety protections for citizens. These are legitimate policy choices, but they must be recognised as distributional trade-offs rather than disguised as pure efficiency gains.

GDPR Reform: Compliance Costs and Market Concentration

The General Data Protection Regulation remains the most prominent target of the deregulatory push. The Commission’s May 2025 Omnibus IV package and the November 2025 Digital Omnibus both propose measures to ease GDPR compliance, particularly for smaller companies. These include simplified consent mechanisms, automated privacy signals, and expanded definitions of what constitutes non-personal data that falls outside GDPR’s scope.

The compliance cost argument for GDPR reform is well-documented. At up to €500,000 per SME, GDPR’s fixed costs create a regressive burden that falls hardest on the firms least able to absorb it. This has contributed to market concentration: larger firms with dedicated privacy teams can comply efficiently, while smaller competitors face proportionally greater obstacles. The result is a regulatory environment that inadvertently favours incumbents — precisely the opposite of the competitive dynamism regulators sought to promote.

However, the Bruegel analysis cautions against treating GDPR reform purely as an efficiency exercise. Research by Aridor et al. (2021) found that implementing an opt-in requirement reduced users of a major online travel intermediary by 12.5%, lowering advertising revenues — a distributional shift from platforms to consumers exercising privacy preferences. Yet the same study found that the remaining users provided higher-quality data because they had actively chosen to share it, potentially improving targeting accuracy and platform value per user. The study by Lefouili et al. (2024) provides further evidence that privacy regulation can actually increase consumer demand and innovation when a significant proportion of users are privacy-conscious — suggesting that some privacy protections generate genuine efficiency gains rather than merely distributing value.

Digital Markets Act and Distributional Consequences

The Digital Markets Act (DMA) represents Europe’s most ambitious structural intervention in digital markets, targeting the largest platform companies designated as “gatekeepers.” The DMA’s provisions illustrate both mitigating and steering distributional effects — a distinction central to the Bruegel analysis.

DMA Article 6(5), which bans self-preferencing by gatekeepers, is an example of a mitigating distributional effect: it redistributes value from dominant platforms to competing businesses and consumers without fundamentally altering business models or technology pathways. When a search engine cannot prioritise its own shopping service over competitors’ results, value shifts from the platform to independent merchants — but the underlying technology and market structure remain broadly similar.

By contrast, DMA Article 6(6), which requires gatekeepers to reduce switching costs and enable data portability, has steering potential: by making it easier for users and businesses to move between platforms, it could alter the fundamental dynamics of digital market competition and potentially redirect technological development toward more interoperable, open standards. The Commission’s approach to enforcing — or relaxing — these provisions carries significant implications for market structure, innovation incentives, and the distribution of value across the digital ecosystem.

AI Act Delays and the EU Digital Deregulation Timeline

The AI Act, Europe’s landmark legislation for governing artificial intelligence systems, has become a key battleground in the deregulatory debate. The November 2025 Digital Omnibus proposed postponing AI Act enforcement from August 2026 to December 2027 — a delay of sixteen months that the Commission justified on grounds of allowing more time for standardisation and industry preparation.

This delay encapsulates the efficiency-distribution tension perfectly. For AI developers and deployers, the postponement reduces near-term compliance costs and allows more time to develop compliant systems — a genuine efficiency benefit. But for citizens and communities exposed to high-risk AI systems — in recruitment, credit scoring, law enforcement, and healthcare — the delay means sixteen additional months without the safety guardrails the AI Act was designed to provide. This is not an efficiency improvement; it is a distributional choice to prioritise industry readiness over citizen protection.

The AI Act also faces challenges around standardisation. The legislation requires high-risk AI systems to meet specific data quality and bias mitigation standards, but these standards remain under development. Article 10(2) on data governance and the obligations to mitigate discriminatory impacts require technical specifications that harmonised European standards bodies have not yet finalised. Research by von Zahn et al. (2022) demonstrates that removing gender bias from e-commerce AI prediction models increased costs for model promoters by 8–10% — quantifying the real economic trade-off between fairness and efficiency that the AI Act attempts to navigate. Understanding the full complexity of AI regulation requires deep engagement with the primary research; exploring interactive analyses of major policy papers can make these trade-offs tangible for decision-makers.

Mitigating vs Steering: Two Types of Distributional Effects

One of the Bruegel brief’s most valuable analytical contributions is the distinction between mitigating distributional effects and steering distributional effects. Mitigating effects redistribute value among existing market actors without changing the underlying business models or technology pathways. The DMA’s ban on self-preferencing is a classic example: it shifts revenue from gatekeepers to competitors but does not alter the fundamental advertising-supported platform model.

Steering effects, by contrast, shape the direction of technological and commercial development. GDPR’s privacy-by-design requirement (Article 25) is a steering intervention: by mandating that privacy be embedded in system architecture from the outset, it potentially redirects innovation toward privacy-preserving technologies such as edge computing, federated learning, and differential privacy techniques. These technologies might not have developed as rapidly — or at all — without regulatory pressure.

This distinction matters enormously for Europe’s strategic autonomy. If the EU merely mitigates distributional harm within existing technology paradigms dominated by US and Chinese firms, it remains a rule-taker in markets shaped by others’ innovations. But if it uses regulation to steer technological development toward pathways aligned with European values and capabilities — privacy-preserving AI, interoperable data infrastructure, trustworthy digital identities — it can create competitive advantages rather than simply managing disadvantages. The risk of the current deregulatory push is that it eliminates steering regulations before their innovation-redirecting potential has been realised.

Evidence on the Real Cost of Digital Fairness

The Bruegel brief assembles empirical evidence that quantifies the trade-offs involved in EU digital regulation. The 12.5% reduction in user participation documented by Aridor et al. (2021) following an opt-in privacy requirement illustrates the immediate revenue cost of privacy protections. The 8–10% increase in model development costs associated with bias removal (von Zahn et al., 2022) quantifies the price of algorithmic fairness. The market concentration increases following GDPR (Peukert et al., 2022) demonstrate the unintended competitive consequences of well-intentioned regulation.

Yet the evidence also reveals countervailing benefits. Privacy regulation can increase consumer trust and demand (Lefouili et al., 2024), potentially generating larger markets that offset compliance costs. Data protection creates incentives for more efficient data use, pushing firms to extract maximum value from consented, high-quality data rather than relying on volume. And standardised regulatory frameworks across the Single Market reduce the fragmentation costs that arise when 27 member states apply different national rules.

The critical point is that these trade-offs exist regardless of whether they are measured. The Commission’s failure to conduct formal impact assessments for many Omnibus proposals means that distributional consequences are being generated without being documented, analysed, or subjected to democratic scrutiny. Citizens are being asked to accept changes to their privacy and safety protections without evidence that the competitiveness gains justify the trade-off — or that the gains will materialise at all.

What Brussels Must Fix: Impact Assessment Failures

The Bruegel brief identifies a fundamental procedural failing at the heart of the deregulatory push: the systematic absence of rigorous impact assessments. Multiple Omnibus proposals have been advanced without the formal ex-ante evaluation processes that the Commission’s own guidelines require. The justification — urgency and the need for speed — undermines the evidence-based policymaking that distinguishes European governance from more arbitrary regulatory approaches.

Without impact assessments, it is impossible to determine whether proposed changes will actually achieve their stated objectives. Will broadening the definition of non-personal data genuinely accelerate AI development, or will it primarily benefit a small number of large data-processing firms? Will easing GDPR requirements for SMEs create competitive dynamism, or will it reduce the privacy protections that European consumers value? Will delaying the AI Act produce better-prepared compliance or simply defer accountability for harmful AI systems? These questions cannot be answered without structured analysis.

The European Ombudsman has raised concerns about the procedural shortcuts taken in Omnibus processes, and the European Parliament has questioned whether bypassing standard legislative scrutiny is compatible with democratic accountability. The Bruegel brief reinforces these concerns with an analytical framework showing that every regulatory change — whether adding or removing rules — involves trade-offs that must be made explicit, measured, and subjected to democratic deliberation rather than presented as technical simplification.

A Roadmap for Transparent EU Digital Deregulation

The Bruegel brief concludes with a set of recommendations that chart a path toward deregulation that is both effective and democratically accountable. First, the Commission must design robust, consistent methodology for impact assessments supporting all deregulatory initiatives — not just major legislative proposals but also the technical amendments and definitional changes that can have profound practical consequences.

Second, Brussels should avoid pre-committing to arbitrary targets for cutting regulatory burden. Setting percentage reduction goals — as some Member States have advocated — creates perverse incentives to eliminate rules based on measurability rather than merit. The objective should be maximising total societal value, not achieving a bureaucratic scorecard target.

Third, every deregulatory measure should explicitly separate and measure three metrics: efficiency effects (changes in total productivity), mitigating distributional effects (redistribution among existing actors), and steering distributional effects (impact on technological pathways and systemic choices). Fourth, the Commission must state explicit ultimate objectives: if the strategy aims to close the EU–US gap, say so openly and discuss the trade-offs involved, rather than presenting distributional choices as mere efficiency improvements.

Fifth, regulatory reform should be combined with complementary policy tools — industrial policy, R&D investment, public procurement standards, and talent attraction — rather than treated as a standalone solution. The competitiveness gap cannot be closed by deregulation alone; it requires positive investment in the capabilities and ecosystems that will drive European innovation. For professionals and organisations working to stay ahead of these regulatory shifts, deep engagement with the primary analysis — through interactive explorations of policy research — provides the foundation for informed strategy in a rapidly changing environment.