ECAT at ECAF'26

ECAF is organised by the European Association for Algorithmic Fairness and seeks to advance the conversation around algorithmic fairness, with a particular focus on innovative research emerging from Europe. The event provides a space for interdisciplinary research dealing with the fields of computer science, law, sociology and philosophy.  

ECAT facilitated an interactive session on data access under the Digital Services Act (DSA). Furthermore, ECAT researcher Erasmo Purificato presented the paper With a Little Help From My Friends: Collective Manipulation in Risk-Controlling Recommender Systems in a spotlight presentation and poster session.

Discussing DSA data access

The ECAT team delivered an interactive session on data access under the DSA in collaboration with Joana Gonçalves de Sá, who is one of the first researchers to apply for vetted researcher status under the DSA. 

In the session, participants were introduced to the data access mechanism, and also gained some insights into the role played by independent research in the enforcement of the DSA. Then Joana shared fascinating findings from her and her colleagues’ research on political bias in search results around European elections, which they are hoping to build on through vetted researcher status under the DSA. 

Participants also shared their own experiences and reflections on data access and the potential utility to their research. This included thoughts on which systemic risks under the DSA (Art. 34) their work relates to, where the most common choice was potential negative effects on fundamental rights (see survey results below).

If you would like to learn more about data access, please see our FAQs.

Beyond data access

The ECAF conference offered a wealth of talks and sessions spanning all corners of the algorithmic fairness discussion. From the impact of AI on democracy to the best technical measures of fairness and explainability to the challenges of meaningful transparency, the three days offered rich conversations across disciplines.

ECAT was also on the programme with the paper With a Little Help From My Friends: Collective Manipulation in Risk-Controlling Recommender Systems, presented by Erasmo Purificato. The paper addresses recently proposed risk-controlling recommender systems that build directly on these feedback signals using binary user input, and how these can be susceptible to coordinated adversarial user behaviour – as well as a potential mitigation strategy. 

Overall, the conference proved highly engaging, and the ECAT team was grateful to have the opportunity to take part and contribute.