Rose Marie Santini, NetLab UFRJ
Hugo Leal, University of Cambridge
Adriano Belisario, NetLab UFRJ
Bruno Mattos, NetLab UFRJ
10.25358/openscience-15829, PDF
The current digital architecture is largely grounded in data regimes, sociotechnical ecosystems that regulate the relationship between digital infrastructures, our online footprint, and private appropriation of public data. Under the current data regime, public social media data is not a freely accessible digital common (Fuster Morell, 2010) but part of broader digital enclosures through which technological corporations extract and fence off commodified data (Andrejevic, 2013). The digital enclosures establish a global hierarchy separating sovereign states that can regulate and enforce new data regimes. This global hierarchy is dependent on the state’s power and corporate influence, and certain states are therefore more vulnerable than others.
The prevailing economic logic of technological corporations led to a continuous deprecation of data access mechanisms. In less than a decade, we witnessed a regressive movement towards data inaccessibility. While previous data regimes contemplated unequal but existing programmatic availability of social media data, the new models significantly foreclosed access. Indeed, after a brief period of “independence by permission” (Wagner, 2023) where platforms dictated which scholars and scholarship could be informed by social media data, the natural corollary of this corporate-controlled data regime materialized as an “APIcalypse” (Bruns, 2019), a deprecation of most forms of programmatic access to data. This represents a threat against scientific research and public oversight.
Meta provides a particularly instructive case for how political communication, platform governance, and regional asymmetries interact. On the user-generated content side, the company’s approach to data transparency illustrates a broader trend from openness to restriction. What began as a relatively permissive developer ecosystem, inviting third parties to build on top of Facebook’s social graph, was progressively curtailed, especially after the Cambridge Analytica scandal heightened concerns about how extensive third-party access to platform data could be exploited, prompting restrictions that ultimately backfired on independent researchers (Venturini & Rogers, 2019). Since then, Meta has moved first toward stricter API permissions and platform review, and later toward increasingly vetted and platform-controlled research tools. Meta’s previous main interface, CrowdTangle, despite its limitations, still offered journalists and academics relatively flexible access to public content and became a key infrastructure for studying political communication on Facebook and Instagram. Its discontinuation in 2024 and replacement by the more restricted Meta Content Library (MCL) marked a further narrowing of independent access. The MCL operates as a controlled-access research environment whose terms limit researchers’ ability to programmatically and independently extract and analyze structured data. Together with X’s closure of academic access and monetization of its API, this transition has been widely criticized as emblematic of the “post-API” era (Tromble, 2021) in social media research.
Data Access Inequalities
Online platforms’ data access regimes appear to be entering a new phase. Technological corporations are adopting a model of selective compliance. According to our findings, while data access is still insufficient, the EU and the UK benefit from more favorable data regimes, particularly when it comes to online advertisement data. This is aggravating the North-South digital divide. For example, as our report also shows, while the EU has established legal mechanisms for researcher access to platform data through Article 40 of the DSA and mandates platform advertising transparency through Article 39, Brazil lacks a comprehensive platform-transparency framework, despite pioneering a civil rights framework for the internet (Moncau & Arguelhes, 2020) and requiring certain data transparency requirements from social media companies for adjacent issues such as women’s rights, child protection, and electoral campaigns.
Empirically documenting and measuring these unequal data access regimes is one of the key contributions of our recent policy report, “Data Not Found” (Santini et al., 2026), which brings results from the Social Media Data Transparency Index, an initiative developed by NetLab UFRJ (Federal University of Rio de Janeiro, Brazil). In its second edition, with the support of the Minderoo Centre for Technology and Democracy (University of Cambridge, UK), the Index systematically reviewed the user-generated content (UGC) and advertising data transparency standards of 15 social media platforms across three jurisdictions: Brazil, the UK, and the EU. The methodology, data, and evaluation sheets are openly available in the project repository[1] to support further research on platform transparency.
While access to social media data has recently deteriorated across the board, the situation is not equally bad. In our assessment, Brazil consistently recorded the lowest overall transparency scores for both public UGC and advertising data, while the EU often achieved the highest. Even in highly regulated jurisdictions, however, the landscape remains highly fragmented and dependent on the platforms’ discretion.
Our findings align with prior evidence that platforms such as X, Snapchat, and Pinterest, despite falling within the scope of the DSA, still fail to meet minimum transparency standards (European Centre for Algorithmic Transparency, 2025). X maintains relatively decent, though costly UGC data access mechanisms, but its advertising repository exists largely in name only in the EU and is unavailable in other jurisdictions. Snapchat and Pinterest provide no functional transparency resource for either content type in any region, except for a narrow set of ad tools nominally available in the EU (and, for Pinterest, also in Brazil), which we also classified as “Negligible” because they were effectively inoperative. Among platforms not covered by the DSA, widely used services such as Discord and Kwai (Kuaishou) offer no data transparency mechanisms at all in any region.
Large companies such as Meta, TikTok, and LinkedIn, in contrast, concentrate meaningful transparency in jurisdictions with stronger oversight, deploying their most functional tools where regulatory pressure is highest and withholding them where it is weakest. TikTok’s Research API, mandated under the DSA in the EU, has also been extended to researchers in the United States and the UK. Although TikTok is not subject to regulatory obligations to provide this API to researchers based in these countries, it has done so, which we hypothesize to be a form of regulatory spillover, part of the “Brussels Effect”, whereby regulatory standards adopted in the EU shape platform practices beyond its jurisdiction (Husovec & Urban, 2024).
Selective Transparency and Political Advertising
Content type also matters across jurisdictions: UGC and advertising data vary significantly by region. This difference has important implications, given that online advertising has been used to enable harmful practices, such as political influence operations and financial fraud. In fact, it is precisely in the domain of advertising transparency, often overlooked, that regional disparities in enforcements become most visible.
The Meta Ad Library reached our highest classification in the UK, where it discloses comprehensive information on all delivered ads and supports structured extraction across different ad types. The same infrastructure is deployed in the European Union, but the bulk-export feature has been disabled from its graphical interface following Meta’s response to the EU on the Transparency and Targeting of Political Advertising Regulation: rather than meeting the regulation’s stricter requirements for political ads, Meta announced in late 2025 that it would stop serving such content in the EU altogether and removed historical political advertising data from its archive (Votta, 2025). Compliance, in this case, produced less transparency than existed before: a paradox researchers had warned about since the earliest generation of platform ad libraries (Leerssen et al., 2019).
In the Global Majority, platforms simply maintain an opaque data regime. In Brazil, the Meta Ad Library score drops to “Deficient”. Structured extraction is restricted to ads self-labeled by advertisers as “political or related to social issues”, a classification over which advertisers have almost full control and can trivially game by reclassifying content as political or not. All remaining ads can only be browsed one by one through the web interface, with no bulk export. The architectural choice matters. Across major platforms, ad repositories require users to search by advertiser name rather than by content, topic, or targeting criteria. If one does not already know who placed an ad, it cannot be found. Coordinated networks operating through dozens of obscure pages, undisclosed third-party funding, or misleading advertiser identities therefore remain structurally undiscoverable, and the search architecture itself becomes a mechanism of opacity (Santini et al., 2026; Zalnieriute, 2021).
The broader implication of treating “political advertising” as a special category seems clear: the boundary is trivially gameable, and trying to police it leaves large blind spots intact (Krekó & Molnár, 2026). The position we defend is therefore straightforward. For large online platforms, all advertising data should be fully and globally accessible, searchable by content, topic, and targeting criteria.
Towards a More Balanced Data Regime
The data regime asymmetries must be addressed through a symmetrical governance framework that enforces universal access to public data. The accelerated development of AI and its equally rapid encroachment in our public and private life is reproducing the same unequal data regimes. Private data extractivism and algorithmic opacity are, once again, taking precedence over public oversight and platform observability (Rieder & Hofmann, 2020). The data regime is very similar, combining advanced forms of digital capitalism (Schiller, 1999) and data colonialism (Couldry & Mejias, 2019) characterized by appropriation and unequal data relations. This is only possible due to an alliance between platform capital and state power that seeks to legitimize the systematic appropriation of our individual and collective footprint through relations that replicate the colonial patterns of extraction and accumulation. Data regimes extend far beyond raw data to include the material infrastructures of digital capitalism, including data centers, cheap energy, and exploitative forms of labor.
The result of selective data access practices is an unequal global regime of platform transparency that produces second-class researchers and citizens across much of the Majority World, leaving them with weaker tools to understand the digital public sphere that increasingly shapes their lives. Citizens and policymakers are directly affected by platforms’ algorithmic distribution, opaque moderation decisions, and advertising microtargeting, yet they lack the systematic data access needed to understand how political discourse circulates, gains visibility (organically or not), and reaches different publics online.
At the intersection of unchecked platform monetization and restricted data access lie online societal harms. Citizens and communities face unprecedented digital threats, from individual abuse to large-scale, coordinated disinformation campaigns, which are becoming increasingly difficult to observe and mitigate. Researchers are denied access to the information required for public-interest scholarship; data-driven investigative journalism is severely constrained, and policymakers are now unable to rely on evidence-based policy. Ultimately, the current data regimes are stripping citizens of their fundamental freedoms and rights, while trying to dismantle any form of democratic accountability. Against the backdrop of data opacity, we propose a new global regime of data governance governed by principles of information integrity and transparency (Santini et al., 2026).
Our report suggests that regulation can push platforms toward greater transparency. That’s why companies face resistance, and even powerful nations and regions struggle to guarantee basic requirements. However, national regulation alone might leave us only with selective transparency decisions. Without clear standards and strong enforcement mechanisms, platforms retain wide discretion over whether, when, and how they comply. When companies implement transparency tools only in response to region-specific regulatory pressure, they reinforce global inequalities in data access. That discretionary power creates and perpetuates large-scale knowledge asymmetries that disproportionately disadvantage those in more vulnerable positions.
Researchers, journalists, regulators, and civil society actors in the most powerful jurisdictions and consumer markets should therefore push for global transparency standards. Without binding, interoperable, and globally applicable data access standards for large online platforms, most researchers will continue to be systematically denied access to the evidence on which knowledge depends. More importantly, these standards need to be developed with researchers, civil society, and regulators in the Majority World as equal stakeholders, not as recipients of whatever spillover the Brussels Effect happens to carry.
References
Andrejevic, M. (2013). Surveillance in the digital enclosure. In The new media of surveillance (pp. 18–40). Routledge. https://doi.org/10.4324/9781315876337-2
Bruns, A. (2019). After the ‘APIcalypse’: Social media platforms and their fight against critical scholarly research. Information, Communication & Society, 22(11), 1544–1566. https://doi.org/10.1080/1369118X.2019.1637447
Couldry, N., & Mejias, U. A. (2019). Data colonialism: Rethinking big data’s relation to the contemporary subject. Television & New Media, 20(4), 336–349. https://doi.org/10.1080/1369118X.2019.1637447
European Centre for Algorithmic Transparency. (2025). FAQs on DSA data access for researchers. European Commission. https://algorithmic-transparency.ec.europa.eu/
Fuster Morell, M. (2010). Governance of online creation communities: Provision of infrastructure for the building of digital commons [Doctoral dissertation, European University Institute]. http://hdl.handle.net/1814/14709
Husovec, M., & Urban, J. (2024, February). Will the DSA have the Brussels Effect? Verfassungsblog. https://verfassungsblog.de/will-the-dsa-have-the-brussels-effect/
Krekó, P., & Molnár, C. (2026). The Houdini ads: How and why political ads are still slipping through the filters of Google and Meta in Hungary. European Digital Media Observatory. https://edmo.eu/blog/the-houdini-ads-how-and-why-political-ads-are-still-slipping-through-the-filters-of-google-and-meta-in-hungary/
Leerssen, P., Ausloos, J., Zarouali, B., Helberger, N., & de Vreese, C. (2019). Platform ad archives: Promises and pitfalls. Internet Policy Review, 8(4). https://doi.org/10.14763/2019.4.1421
Moncau, L. F. M., & Arguelhes, D. W. (2020). The Marco Civil da Internet and digital constitutionalism. In G. Frosio (Ed.), The Oxford Handbook of Online Intermediary Liability. Oxford University Press.
O’Reilly, T., Strauss, I., & Mazzucato, M. (2024). Algorithmic attention rents: A theory of digital platform market power. Data & Policy, 6, e6. https://doi.org/10.1017/dap.2024.1
Rieder, B., & Hofmann, J. (2020). Towards platform observability. Internet Policy Review, 9(4). https://policyreview.info/articles/analysis/towards-platform-observability
Santini, R. M., Leal, H., Salles, D., Belisário, A., Mattos, B., & Pinho, D. (2026). Data not found: Transparência de redes sociais para a integridade da informação. NetLab UFRJ & Minderoo Centre for Technology and Democracy. https://doi.org/10.5281/zenodo.20084860
Schiller, D. (1999). Digital capitalism: Networking the global market system. MIT Press.
Tromble, R. (2021). Where Have All the Data Gone? A Critical Reflection on Academic Digital Research in the Post-API Age. Social Media + Society, 7(1), 2056305121988929. https://doi.org/10.1177/2056305121988929
Venturini, T., & Rogers, R. (2021). “API-Based Research” or How can Digital Sociology and Journalism Studies Learn from the Facebook and Cambridge Analytica Data Breach. Digital Journalism, 7(4), 532–540. https://doi.org/10.1080/21670811.2019.1591927
Votta, F. (2025). What data reveals about Meta and Google’s political ad ban in the EU. Tech Policy Press.
Wagner, M. W. (2023). Independence by permission. Science, 381(6656), 388–391. https://doi.org/10.1126/science.adi2430
Zalnieriute, M. (2021). Transparency washing in the digital age. AJIL Unbound, 115, 9–14. https://doi.org/10.1017/aju.2020.92
Rose Marie Santini is an Associate Professor at the School of Communication of the Federal University of Rio de Janeiro (UFRJ). She is the founder and director of the NetLab – Laboratory of Internet Studies and Social Media at the Federal University of Rio de Janeiro, Brazil. NetLab is a research laboratory at the School of Communication, Federal University of Rio de Janeiro, with an agenda for the diagnosis of the digital disinformation phenomenon and its societal consequences in Brazil. NetLab develops computational, digital and social science methods to expand, create and implement research strategies to empirically and critically investigate the effects of the media ecosystem on public opinion – including mass media, alternative media, hyperpartisan media, junk news sites and social media analysis. NetLab’s mission is to produce empirical evidence to advance scientific knowledge, ground public debate and inform the development and governance of new technologies in the face of contemporary media manipulation strategies. Marie Santini is also a research member of the European VOX-Pol Network of Excellence; a member of the International Observatory on Information and Democracy (OID) and a research member of the Scientific Committee of the IPIE (International Panel on the Information Environment).
Hugo Leal is an Associate Teaching Professor in Digital Humanities (DH), University of Cambridge. He is the Director of Postgraduate Studies in DH at the Faculty of English/Cambridge Digital Humanities (CDH). He is also Research Associate at the Minderoo Centre for Technology and Democracy (MCTD), University of Cambridge. His research focuses on collective action, mis/disinformation, internet cultures and subcultures, and the lifecycle of viral narratives, combining sociology, political science, network theory, and digital methods. He was co-investigator in the EU Horizon-Research and Innovation Action project, “AI4Trust: AI-based Technologies for Trustworthy Solutions Against Disinformation”. He holds a PhD in Political and Social Sciences from the European University Institute (EUI), Florence.
Adriano Belisario is a PhD candidate at Federal University of Rio de Janeiro and assistant researcher at NetLab UFRJ. He holds an MSc in Social Data Science from the Oxford Internet Institute (University of Oxford) and a M.A. degree in Communication and Culture from UFRJ. Belisario has taught data journalism and OSINT (open-source intelligence) techniques in postgraduate and online courses. Belisario has several publications on digital media, communication, and society in Brazilian and top-notch international journals.
Bruno Mattos is a journalist and a projects coordinator at NetLab UFRJ. He earned a Master’s degree in Information Science from the Federal University of Rio de Janeiro (UFRJ). His work examines issues related to digital governance, platform regulation, and content moderation, with an emphasis on the challenges and developments of the Brazilian platform ecosystem.