ABSTRACT

Artificial intelligence is increasingly integrated into spaces of art and culture, promising efficiency, personalisation, and expanded access. Yet these systems operate within structures shaped by political austerity, privatisation, and global inequalities.

Curating Bias examines how generative AI reshapes curatorial practice, alters exhibition narratives, and reinforces existing power relations through opaque decision-making, hidden labour, and algorithmic feedback loops. By tracing the social, economic, and environmental conditions that underpin AI, the book argues that technological innovation in the cultural sector cannot be understood apart from the structural biases it reproduces.

Curating Bias calls for critical transparency and renewed public responsibility in the design and governance of cultural spaces, positioning generative AI as a form of speculative practice that opens exhibition design to critical imagination and shared reflection on how knowledge is visualised in contemporary design.

Artificial intelligence and algorithms are inevitably becoming part of exhibition design. They shape how cultural experiences are imagined, organised, and shared. Their presence is not in itself a problem. Many of these systems are introduced with the aspiration of making exhibitions more accessible, efficient, or responsive to audiences. Yet good intentions are not a safeguard against harm. When designers and curators rely on algorithmic tools, they also inherit the responsibility to question how these tools work and to make their inner logic visible. The data that drives design decisions is never neutral, it is shaped by the values and priorities of those who collect, interpret, and prioritise it. Every dataset reflects a view of the world, a judgment about what deserves attention and what can remain unseen. These are not just technical or logistical decisions but moral ones that influence how culture and history are framed and remembered.

Exhibition design has always been a practice of interpretation, of arranging space and narrative so that meaning can emerge. In the age of AI and rapidly evolving algorithms, this task also involves understanding the systems that now increasingly shape curatorial meaning. Transparency and accountability become part of the design language itself, not external requirements but essential conditions for trust.

To design with algorithms is also to design with awarness of their limits, their blind spots, and their histories. Bias may never be fully removed, but it can be recognised, contextualised, and discussed. When curators and designers take this work seriously, exhibitions can become places that not only show objects or stories but also reflect on the structures that shape how we see them.

As AI becomes increasingly embedded in exhibition design, its influence extends far beyond technical efficiency. The integration of generative and corrective models into curatorial workflows reshapes how institutions produce and display knowledge. What appears as enhancement or restoration can unintentionally act as a form of selection and therefore exclusion. A process through which certain narratives are highlighted, while others get overlooked. When unlabelled, these algorithmic manipulations risk maintaining the aesthetic and ideological assumptions of the global North by homogenising cultural content. To curate in the age of AI is therefore to curate its politics. Museums and exhibition designers should confront the algorithm not as a neutral collaborator but as a participant in the construction of historical truth. Transparency in the use of AI cannot remain a technical question of watermarking or metadata alone. It must evolve into an ethical and curatorial principle. Each alteration, reconstruction, or generated image should be treated as a site of negotiation, between evidence and imagination, preservation and erasure, past and possibility. By labelling AI-based alterations and adopting a speculative approache, curators can transform AI from an instrument of optimisation into a medium of reflection. Speculative imagery reveals rather than conceals manipulation, inviting audiences to engage critically with the processes that shape what they see. Through transparent labelling and spatial differentiation, exhibitions can reclaim their role as spaces of exploration rather than illusion.

In the end, the challenge is not merely to display AI-generated images responsibly, but to reimagine the exhibition itself as a platform for algorithmic awareness. By exposing how technologies participate in shaping cultural memory, institutions can restore curatorial practice. One that keeps culture and history visible, accountable, and alive.

→ Excerpts from Curating Bias.

PORTFOLIO 

Find my CV as well as my Portfolio on my Website.