And the lucky last speaker in the final session at the Social Media Access Days at the German National Library in Frankfurt is the excellent Giada Marino, whose interest is the operation of data access provisions under DSA Article 40.12. The focus here is on very large online marketplace platforms and the systemic risks they pose to minors.
Article 40.12 addresses access to publicly available data from Very Large Online Platforms and Search Engines; this is coordinated across EU jurisdictions by the Expert Group on Access to Publicly Available Data (ECAT). The present project focusses on marketplace platforms such as Amazon, Facebook Marketplace, Temu, AliExpress, and Shein, and made data access requests to these platforms under DSA Article 40.12.
This therefore covers both business-to-consumer and consumer-to-consumer platforms, and explores whether such platforms might advertise illegal content such as unsafe products or deceptively listed materials, and how they protect minors from age-restricted material or harmful products. The focus here was especially on e-cigarettes, toys and games, cosmetics, and girls’ clothing, as potentially sensitive product categories.
For these, the project gathered product-level data (including product information, visibility metrics, product labels and warnings, and age restrictions), seller characteristics (including seller IDs, country of origin, and other metadata), user-generated content (reviews and ratings, seller interactions, user comments, and associated visual media), and preferred data format information (CSVs, identifiers, APIs, JSON, etc.).
AliExpress provided API access within two days; Meta took two months and provided only generic Meta Content Library tools; Temu took 78 days and offered structured, tabular data; Shein took 101 days, and imposed various legalistic barriers, but also offered structured, tabular data; and Amazon took even longer, and strangely simply whitelisted the research team to enable them to scrape the Amazon Website (the usability of this approach has yet to be established).
What emerged from this is that Amazon and Meta did not have specific channels to request data access under Article 40.12; researchers had to rely on genetic communication pathways which are poorly documented. Most platforms took a substantial amount of time to respond to access requests, and Shein in particular imposed significant legal barriers to data access and use; data access and completeness across platforms also varied widely. This led to researchers having to adjust their original research plans, in order to harmonise the datapoints and timeframes across all platforms.
There is therefore a pressing need for regulators to mandate more structured data dictionaries for DSA requests; they should also enforce legally binding timelines for request responses, since most research projects are time-limited and lengthy delays in data access result in a severe reduction in the time available for research; and they should require the establishment of dedicated communication channels for DSA Article 40.12 requests.
