This publication examines how Small and Medium-sized Enterprises (SMEs) perceive and address gender bias in artificial intelligence systems.
While AI bias is increasingly recognised as a critical concern, existing literature predominantly focuses on large technology firms, leaving a significant gap regarding the experiences of SMEs. This research addresses that deficiency by investigating the specific structural constraints and organisational realities that shape how European digital SMEs engage with AI fairness.
Drawing on 39 semi-structured interviews, the analysis reveals that while awareness of gender bias risks is generally high, SMEs predominantly frame the issue as a technical challenge rooted in data engineering rather than a sociotechnical one. Consequently, mitigation strategies are largely confined to technical safeguards, such as dataset partitioning and quality control, rather than inclusive design practices.
European SMEs operate within a fragmented ecosystem marked by limited agency and infrastructural dependency on external providers. To move beyond data-driven fixes toward holistic ethical-by-design approaches, SMEs require targeted policy support.
This publication can be downloaded from Zenodo. The publication hasn’t yet been reviewed and approved by the European Commission.
Authors: Alexandros Minotakis, Elizabeth Farries, Loredana Bucseneanu, Sandra Sieron, and Molly Newell.
