Executive summary
This paper investigates how the European Commission’s Better Regulation Framework (BRF) was applied in the development of three landmark digital law initiatives: the Artificial Intelligence Act (AI Act), the Digital Markets Act (DMA) and the now‑withdrawn proposal for a Standard Essential Patents (SEP) Regulation. The BRF is a set of guiding principles, tools and procedures established in 2015 by the EU Commission to promote targeted, effective and minimally burdensome EU regulation. The paper analyses the process followed by the EU Commission in elaborating three legislative proposals in light of BRF core stages, namely the Stakeholders Consultation, the Impact Assessment and the Explanatory Memorandum attached to the legislative proposal. The conclusions and the final evaluation table summarise the main comparative findings of the paper.
The DMA, the AI Act, and the draft SEP Regulation were proposed during the first Von der Leyen Commission. They represent the EU’s attempt to regulate the digital economy, trying to safeguard fundamental rights, competition and innovation. Secondly, the legislations all rely on the same legal basis – i.e. Art. 114 Treaty on the Functioning of the European Union (TFEU), the EU internal market legal basis. Since the EU Commission proposed the legislation after the entry into force of the BRF framework in 2015, they represent good case studies to check the EU Commission's compliance with Better Regulation practices.
Unlike other works that extensively analyse the ‘content’ of these Regulations, this study focuses on the ‘legislative steps’ followed by the European Commission in elaborating the legislative proposals of the AI Act, DMA and SEP Regulation in the light of the BRF. Similarly, the study does not analyse the legislative steps followed by the European Parliament and the Council in adopting these legislations, since these EU institutions have never adopted a framework of legislative best practices comparable to the BRF.
The Commission launched multiple public consultations on the AI Act, first in response to the 2020 White Paper on AI and then during the 2021 inception phase. These exercises generated a wealth of stakeholders’ input, yet the feedback remained dispersed across annexes and online repositories without a consolidated synopsis report accompanying the legislative proposal. The Regulatory Scrutiny Board (RSB) initially criticised the AI Act Impact Assessment for methodological weaknesses, particularly concerning small and medium‑sized enterprises (SMEs) cost estimates. Although a revised assessment ultimately satisfied the RSB’s procedural requirements, only partial adjustments were made to its recommendations. The Explanatory Memorandum accompanying the AI Act proposal is formally comprehensive in detailing the legal basis, subsidiarity and proportionality considerations, and procedural history. However, it does not summarise experts’ contributions or present the full content of RSB opinions, leaving gaps in transparency regarding how external advice shaped the final text of the AI Act proposal.
In June 2020, the EU Commission organised two inception‑phase consultations on a New Competition Tool (NCT) and a Digital Services package proposal, which later evolved into the DMA and the Digital Services Act (DSA). Despite the public consultations, the EU Commission did not issue a synopsis report explaining why significant suggestions expressed by the stakeholders were not taken on board. In particular, the EU Commission never explained the reasons for abandoning the NCT proposal and opting for the DMA. In addition, contrary to the BRF framework, the EU Commission did not organise any public consultation on the final version of the DMA proposal after it was published in December 2020. The RSB approved the DMA proposal “subject to reservations” concerning evidence strength and benefits estimation, and several of its concerns remain visible in the published documentation. In particular, the RSB noticed that the DMA Impact Assessment relied on anecdotal evidence to identify “core platform services” subject to the scope of DMA and the effects of DMA on innovation. The DMA Impact Assessment presented three policy options (i.e., from a closed list of self-executing obligations to a fully dynamic regime) and ultimately endorsed a semi‑dynamic approach. Contrary to the BRF, quantified benefit and cost estimates were provided only for this middle‑path option, which had been pre-selected by the EU Commission.
The EU Commission’s proposal for an SEP Regulation never progressed to adoption, yet its preparatory work illustrates challenges faced by the EU Commission in self-complying with the BRF framework. Public and SME-specific consultations revealed polarised positions: implementers pressed for greater licensing transparency and mandatory essentiality verification, while SEP holders supported transparency but raised several procedural and substantial concerns including the lack of empirical support and the risk of over-regulation. Despite drawing on specialised studies, the Impact Assessment remained theoretical and insufficiently anchored in empirical data; it also overlooked the implications of Europe’s forthcoming Unitary Patent system. The Explanatory Memorandum attached to the SEP Regulation proposal complied with formal structuring, but failed to deliver a detailed proportionality analysis or to synthesise expert or RSB viewpoints. Finally, the EU Commission decision in February 2025 to withdraw the proposal “for lack of foreseeable political agreement” was announced without the kind of documented rationale the BRF purports to require.
Across all three initiatives, the EU Commission partly adhered to BRF formalities yet exhibited inconsistent application of its best‑practice commitments. First, stakeholder feedback was rarely consolidated into accessible synopsis reports, empirical evidence remained uneven, and transparency regarding expert input and feedback from the Regulatory Scrutiny Board was limited. To strengthen the BRF’s effectiveness, the paper recommends mandating unified consultation synopses, requiring breakdowns of costs and benefits by stakeholder group, and extending binding BRF obligations to the Council and the European Parliament. Finally, in light of the lack of transparency followed by the EU Commission in withdrawing the SEP Regulation proposal in February 2025, the paper proposes codifying the withdrawal rationale into the BRF framework. These reforms would bolster EU legislation’s clarity, legitimacy and resilience in the face of fast‑moving technological and market developments.
The paper's findings are relevant in light of the ongoing policy debate in Brussels on the effectiveness and abundance of the EU digital acquis following the criticism expressed by the Draghi report. The paper argues that, if the EU intends to regulate the digital economy further, it should follow a straightforward evidence-based approach, as indicated in the Better Regulation Framework.
