
Visual Imaginaries of Gender: The potentials and limitations of GenAI
This project investigates the gender imaginaries of Generative Artificial Intelligence (GenAI) for image generation. Whilst prior research has demonstrated that GenAI defaults to masculine when prompted without gender information, the full spectrum of GenAI’s artistic reimaginations of gender remain unexplored.

This project will investigate this with a mixed-methods approach, combining technical experimentation, perception studies, and historical contextualisation. In collaboration with artists, the researchers will conduct workshops to collect data and gain critical insights in the use of GenAI. Through this research, they seek to map GenAI's imaginaries of gender and contribute to the development of more inclusive, creative tools for research, teaching, and artistic practices.
The project’s innovation lies in its shift from studying default biases to exploring the full range of GenAI’s artistic reimaginations of gender across cultural and historical dimensions. It will examine both limitations and potentials of these systems, aiming to overcome constraints and unlock new creative possibilities. With this project, the researchers seek to deepen understanding of how AI technologies influence perceptions and representations of gender.
The project's partners, Sandberg Instituut, the Hogeschool of Amsterdam, and Studio Bertels, will contribute artistic and creative industries expertise, grounding the project within the communities most affected by GenAI.
Project team:
- Nanne van Noord (FNWI)
- Eftychia Stamkou (FMG)
- Melvin Wevers (FGW)
- Sabine Niederer (HvA)
- Maarten Groen (HvA)
- Flavia Dzodan (Sandberg Instituut)
- Dirk Bertels (Studio Bertels)
Now that anyone can create images with the aid of AI, we are seeing more and more AI-generated images around us. At the same time, little is known about how the underlying AI systems work and how biased they are – for ex...