Participate in the DSA Ad Repository Data Sprint

Context

The Digital Services Act (DSA) was formally adopted in July 2022 and came into action on 17 February 2024. Among the various transparency measures introduced by the DSA, Very Large Online Platforms (VLOPs) have been obligated to create and maintain a Public Ad Repository that includes the content of each paid advertisement (text and image), information about the advertiser, typically including their identity, data regarding the target audience (demographic or interest-based parameters). The goal is to provide transparency that enables independent scrutiny and research regarding the influence of advertisements, particularly those that could impact democratic processes, public health, or societal debates. These repositories must be easily accessible for users, researchers, and regulators to scrutinize the nature and scope of advertising activities on the platform.

Nearly one year after the DSA has entered into force, the time has come to reflect on the real use of these ad repositories. The Open Institute for Digital Transformations, created as a part of TIERED, proposes to do this via a data sprint tailored to enhance collaboration and knowledge sharing among academic scholars, regulators, journalists and activists. This collective experiment will allow these communities to explore various risks linked to online advertisements: misinformation, undeclared political discourse that may interfere with election processes, unethical corporate strategies, health-related campaigns, etc. Most VLOPS have established a public API which can be accessed by researchers, journalists or activists: among others, Facebook has extended its Ad Library and Tiktok is offering an API access to its Commercial Content API, while Google has maintained its Google Ads Transparency Center allowing access to information about content that was moderated by the search engine.

Format

We invite a wide array of experts from regulatory agencies and civil society organizations, data journalists, data scientists, social media researchers, legal scholars, sociologists, economists, computer scientists to work together with Sciences Po students in a Data Sprint to explore the possibilities offered by these repositories. During the sprint, teams consisting of students, engineers, and researchers from Sciences Po and interested partners will engage in an intensive program from February to April.

The “sprint” approach is aimed at preventing participants from becoming mired in exhaustive analysis, allowing them to focus on innovation and practical solutions. Teams will be guided by experts and engineers in using appropriate data sources, and data analysis techniques to answer a specific research question. The overarching goal is to bolster the collaboration between research bodies, civil society and regulatory agencies, thereby creating an optimal environment for collaboratively developing potential policy recommendations for the successful rollout of the DSA.

Around five to ten teams will participate in three structured collaborative work sessions. The teams will be composed of the following members:

• Experts from academia, regulatory organizations, civil society, etc. with a familiarity with the substantive questions posed by the DSA / Ad repository
• Engineers / data scientists, who will be able to collect and analyze the data
• Students who are curious about the sprint, and who have good analytical skills and a basic knowledge of computational approaches"

Group leaders will be identified and selected in advance based on their interest and experience with the ad repository data. Masters’ level students from Sciences Po will be selected, alongside the team leaders, based on their interest, skills, and motivation.

The sprint will be split into three phases, designed with distinct objectives and will include various group activities that will take place between February and April. Before the start of the first phase, group leaders are being identified and pre-work the data to determine feasibility of certain research questions. Teams will be formed during the first phase.

• Phase 1 - Project Design (1 full day, 24 February): Presentation of the objectives of the Data Sprint. The morning session will feature a certain number of keynote speakers: presentation of the DSA, specifically Article 39, and its challenges by Sciences Po researchers (ex. Beatriz Botero Arcila) and a regulator (ex. Arcom). Team Leaders will pitch their research questions and students will choose which group they want to join. We will make sure all teams will gather complementary expertise. In the afternoon, all groups will meet to discuss shared tasks and challenges as well as to establish guidelines to continue working remotely in their groups (selecting/acquiring the relevant dataset and skills, reviewing the literature, etc.).
• Phase 2 - Data-intensive Sprint (3 afternoon sessions, 25, 26, 27 February): In this stage, teams will gather to work on their dataset, addressing both informational and technical issues that arise during their analysis. Guidelines will be provided to structure the progress of each team and facilitate knowledge and skill transfer.
• Phase 3 - Results and Recommendations (1 afternoon, 15 April): The final phase involves teams presenting their results to an audience composed of researchers, policymakers, and regulators. Concluding remarks about collective recommendations for the ad repository reports will be produced, possibly through a collective policy brief.

Between each phase, groups are expected to work together virtually.

Participate in the data sprint

If you are interested in participating in the data sprint, either as a researcher, student, engineer, regulator, or representative of civil society, we invite you to fill out this form. 

The deadline for submissions is 30 January. 

For any questions, please contact: carly.hafner@sciencespo.fr.