114 Nexa Lunch Seminar - copertina
Mer 22 Mag
Seminari e Convegni

Personal Data Production: where AI labour & GDPR regulation intersect

With the development of AI systems, that require immense amounts of data, personal data regulation is of strategic importance: considered as a legal matter as much as an economical field, different stakeholders are engaging actively in sometimes conflicting interpretations of what the law says, or what it should say.
An established corpus of law studies has extensively developed the matter of privacy and its translation into law as the protection of "personal data". Its history in the European Union is one of convergence and harmonization between all member states' legal traditions, culminating into the General Data Protection Regulation of 2016.
Another domain of research, invested by sociologists and social scientists, is that of digital labor: the production and annotation of data for AI training and correction. Work platforms, subcontractors, but also unknowing users are participating actively in all sorts of tasks that are essential to the production of AI systems. Companies such as Google, OpenAI and Meta openly admit to scraping texts, photos and videos off the internet, transforming, a posteriori, information relevant to personal data protection into personal data production.
This presentation will give a brief oversight of a different definition of regulation, understood not as what the law says, but as the resulting effective pact between all stakeholders that interpret the law and apply it a certain way. In this perspective, the actors that are central to the personal data regulation, and thus AI production in the EU, are Data Protection Authorities. There are 27 national authorities, and two european agencies that centralize them. The difficult process of agreeing on the application of the GDPR is prolonged into the sphere of AI production: how do they acknowledge the perspective of personal data protection in AI production?

Speaker: Thomas Le Bonniec - Institut Polytechnique de Paris

Thomas Le Bonniec is a Sociology PhD researcher at the Institut Polytechnique under Antonio Casilli and Corinne Vercher's supervision, member of the Digital Platforms Laboratory (DiPLab). Working on the recognition of AI workers by data protection regulation and digital commons. Obtained a master's degree in sociology in 2018 at Sorbonne-Université. In 2019, helped uncover how Siri, Apple's vocal assistant, collected users' audio recordings without their knowledge. Has worked on the digital industry from a critical perspective ever since. Participates in advising public decisionmakers, civil society NGOs and trade unions. Also writes news articles on surveillance, platforms, digital labor, the environmental cost of digital economies and regulation.