[Speech] Antonio Casilli at the Swiss Sociological Association’s conference

On September 4, Antonio Casilli delivered the closing keynote at the Swiss Sociological Association Congress, which took place at the University of Fribourg and the Haute École de Travail Social de Fribourg, around the 2026 theme “Communities & Societies in Movement.”

His keynote was titled “Qui produit l’intelligence artificielle ? La création d’une classe ouvrière des données” (“Who Produces Artificial Intelligence? The Emergence of a Data Working Class”).

In his keynote, Casilli discussed the emergence of two social classes around data work. The first, which he describes as the “digital middle class,” consists of qualified data workers, such as data analysts, engineers, and machine learning researchers.

The second category consists of a much larger, yet highly invisibilised, workforce of data workers. These workers perform tasks such as data annotation, content moderation, and rating or evaluating AI-generated responses. Many of these workers are concentrated in developing countries and occupy lower positions in the AI supply chain.

Casilli argues that the apparent divide between these two groups is not as deep as it may seem. His central argument is that the supposed divide between qualified workers in the technology sector and supposedly unqualified data workers is more closely linked to differences in social and political positioning than to differences in skills or competencies.

A second key argument concerns the consequences of this distinction for solidarity among workers across the AI supply chain. By framing some workers as “qualified” and others as “unqualified,” this division contributes to a broader separation between “digital middle class” in the Global North and data workers in the Global South.

His keynote also highlighted the transnational conditions that could foster the emergence of global solidarity among workers across the AI data supply chain.