reflections from the primary report underneath Article 35(2) of the Digital Providers Act – Official Weblog of UNIO – Cyber Tech

Jamily Maiara Menegatti Oliveira (Masters in European Union Regulation from the College of Regulation of College of Minho)

As we talked about in final week’s put up, on 18 November 2025, the European Board for Digital Providers – in cooperation with the European Fee – revealed the primary annual report underneath Article 35(2) of the Digital Providers Act (DSA), devoted to figuring out essentially the most distinguished and recurring facets of systemic dangers related to Very Massive On-line Platforms (VLOPs), in addition to the respective mitigation measures.^[1] That is an institutionally related doc, not solely as a result of it inaugurates a brand new reporting cycle underneath the DSA, however above all as a result of it reveals how the European Union is starting to translate traditional democratic considerations – such because the integrity of public debate and electoral processes – into the language of danger governance.

Among the many varied areas analysed, Part 3.3 stands out, through which the report identifies systemic dangers related to civic discourse, electoral processes and public security. The choice to deal with these parts as structural dangers, quite than as remoted incidents or mere pathologies of the digital atmosphere, is especially important. It highlights the popularity that the common functioning of platforms – via their algorithmic methods, amplification fashions and moderation practices – can profoundly have an effect on the situations underneath which public opinion is fashioned in a democratic society.

Among the many most recurrent dangers, the report highlights the systemic dissemination of disinformation and deceptive info relating to election dates, candidate eligibility, registration or voting procedures, in addition to the circulation of narratives aimed toward delegitimising election outcomes via unfounded allegations of fraud or institutional interference. These phenomena should not analysed as remoted incidents, however as predictable results of data ecosystems designed to maximise attain and engagement, usually on the expense of data reliability.^[2]

It ought to be famous that the DSA doesn’t regulate elections or electoral processes as such however recognises that the structural functioning of digital platforms can generate systemic dangers for electoral processes by affecting civic discourse, entry to info and the integrity of public debate.

Certainly, algorithmic advice raises a number of considerations, such because the potential polarising impact and the flexibility to create “info bubbles”, in addition to the danger of abuse of economic and political energy and the possibly manipulative impact that suggestions can have on customers’ privateness, autonomy and self-determination.^[3]

Though the present ease of entry to numerous and pluralistic info on digital platforms could make the democratic course of extra participatory and inclusive,^[4]new applied sciences additionally allow the dissemination of misinformation on a big scale, with unprecedented velocity and concentrating on accuracy. As info flows change into more and more personalised, echo chambers and the dynamics of reinforcing pre-existing beliefs additionally intensify,^[5] making residents significantly weak to the affect exerted by folks, information, States, organisations, software program and units, in addition to their respective programmers and regulators.^[6]

Digital platforms are presently extraordinarily essential instruments for the train of democracy, not solely due to the benefit of interplay between residents, representatives and candidates (or would-be candidates), but in addition due to their dynamic nature and the benefit with which content material might be disseminated. Nevertheless, this stream of data can serve a goal reverse to this aim, proving to be a fertile floor for disinformation. Within the electoral course of, disputes are fashioned via the dialectic between candidates, in order that disinformation on this space is much more worrying, since info might be accessed immediately and the unfold of disinformation throughout an electoral course of can considerably affect the electoral will, as sharing is often executed with out checking the veracity and reliability of the data.^[7]

Disinformation can goal the integrity of the electoral course of by discrediting the voting, vote counting system, in addition to candidates or political events, with the intention of gaining a bonus for the opposition.^[8]

In conventional media, throughout election durations, rules are often put in place to make sure that voters have entry to a plurality of voices. Nevertheless, on-line platforms should not topic to the identical ranges of regulation, making them extra prone to disinformation and on-line micro-targeting of voters, a political observe used to govern public opinion via knowledge evaluation and on-line content material concentrating on.^[9]

Explicit concern is expressed concerning the impression of generative synthetic intelligence on electoral processes. The report highlights dangers related to using chatbots and computerized content material technology methods that may present incorrect details about elections or candidates, in addition to the circulation of artificial content material designed to imitate dependable sources or public figures.^[10] In electoral contexts, characterised by excessive temporal and emotional sensitivity, these instruments allow subtle types of info manipulation, posing new challenges to the safety of democratic integrity.

On this context, additionally it is essential to say the contribution of Regulation (EU) 2024/1689, which establishes harmonised guidelines on synthetic intelligence (AI Act), and strengthens the safety of democratic processes within the digital atmosphere. Below Article 6(2), together with Annex III, AI methods designed to affect the result of an election or referendum, or the electoral behaviour of pure individuals within the train of their proper to vote, are categorised as high-risk AI methods. This regulatory classification displays the express recognition that sure functions of synthetic intelligence pose structural dangers to the integrity of democratic processes, justifying the imposition of enhanced obligations by way of design, use and supervision. The hyperlink between the DSA and the AI Act thus reveals a complementary European method: whereas the DSA focuses on figuring out and mitigating systemic dangers related to the functioning of digital platforms, the AI Act intervenes straight on particular applied sciences that would severely compromise residents’ political self-determination.

A further ingredient value highlighting is the time dimension of electoral processes, which poses particular challenges to the risk-based governance method adopted by the DSA. Election campaigns are characterised by quick durations, excessive communication depth and fast circulation of content material, which limits the effectiveness of mitigation mechanisms which can be overly depending on ex put up assessments or gradual compliance procedures. On this context, the flexibility of platforms and competent authorities to establish and mitigate dangers in a well timed method is especially essential, in any other case regulatory responses could show to be structurally out of step with the velocity of data manipulation throughout important moments in democratic life.

Though the report doesn’t expressly check with the idea of the rule of legislation, the centrality given to civic discourse and the integrity of electoral processes reveals a fabric concern with the democratic high quality of the digital public area. By framing disinformation, polarisation, info manipulation and intimidation of actors within the public debate as systemic dangers, the Council and the Fee assume that the safety of democracy within the European Union relies upon, to a big extent, on the preservation of open, pluralistic and safe info ecosystems.

The evaluation of the electoral facet of the primary report adopted underneath Article 35(2) of the DSA reveals a big evolution within the European method to defending democracy within the digital atmosphere. By recognising that the dangers to electoral processes stem largely from the structural functioning of digital platforms and their amplification and moderation methods, the European Union is shifting away from purely reactive responses and in direction of a preventive mannequin of danger governance. Nevertheless, the effectiveness of this method will depend upon the flexibility to make sure swift, clear and proportionate enforcement of the obligations imposed on platforms, particularly throughout election durations marked by excessive informational sensitivity. Electoral integrity is thus established as one of many principal standards for assessing European regulatory ambition and its capability to safeguard a very pluralistic and democratic digital public area.

[1] European Board for Media Providers, First report of the European Board for Digital Providers in cooperation with the Fee pursuant to Article 35(2) DSA on essentially the most distinguished and recurrent systemic dangers in addition to mitigation measures, 18 November 2025, 48, https://digital-strategy.ec.europa.eu/en/information/press-statement-european-board-digital-services-following-its-Sixteenth-meeting.

[2] European Board for Media Providers, First report of the European Board for Digital Providers in cooperation with the Fee pursuant to Article 35(2) DSA on essentially the most distinguished and recurrent systemic dangers in addition to mitigation measures, 20.

[3] See Natali Helberger et. al, “Regulation of stories recommenders within the Digital Providers Act: empowering David towards the Very Massive On-line Goliath”, Web Coverage Assessment, 26 February 2021,  https://policyreview.information/articles/information/regulation-news-recommenders-digital-services-act-empowering-david-against-very-large.

[4] See Felisbela Lopes, “Os desafios do pluralismo e da liberdade dos média diante  da transição digital: o caso do Serviço Público de Média’, in E-E-book CitDig: Centro de Excelência Jean Monnet Em Cidadania Digital & Sustentabilidade Tecnológica – II Congresso Ibero-Americano Sobre Direito e Tecnologias Digitais Subordinado Ao Tema “Estado de Direito e Transição Digital, ed. Alessandra Silveira and Maria Inês Costa (Pensamento Sábio – Associação para o conhecimento e inovação, 2025), 17, https://doi.org/10.21814/1822.97169.  

[5] European Fee, Communication from the Fee to the European Parliament, the Council, the European Financial and Social Committee and the Committee of the Areas, Tackling on-line disinformation: a European Strategy, Brussels, 26.4.2018, COM(2018) 236 closing, https://eur-lex.europa.eu/legal-content/EN/TXT/?uri=CELEX:52018DC0236.

[6] See Jonathan Bridenbaker, “The digital citizen as technoliberal topic: the politics of constitutive rhetoric within the European Union’s digital decade“, Communication and Democracy 58, no. 2 (2024): 163,  https://doi.org/10.1080/27671127.2024.2385912.

[7] See Rodrigo López Zilio, Direito Eleitoral (Editora Juspodivm, 2023), 504–5.

[8] Zilio, Direito Eleitoral, 506.

[9] Elda Brogi et al., “EU and media coverage: conceptualising media pluralism within the period of on-line platforms. The expertise of the Media Pluralism Monitor”, in Analysis Handbook on EU Media Regulation and Coverage, ed. Pier L. Parcu and Elda Brogi (Edward Elgar Publishing, 2021), 28,  https://doi.org/10.4337/9781786439338.00007.

[10] European Board for Media Providers, First report of the European Board for Digital Providers in cooperation with the Fee pursuant to Article 35(2) DSA on essentially the most distinguished and recurrent systemic dangers in addition to mitigation measures, 20.

Image credit score: by Element5 Digital on pexels.com.