Artificial Intelligence (AI) tools apparently offer simple solutions to the challenges facing the modern museum. However, in a world of decolonisation, can a museum ever use AI ethically? This question was at the heart of one of the projects funded by the UK government’s Bridging Responsible AI Divides (BRAID) programme, which focused on the Indian material held by the Royal Armouries, the UK’s national collection of arms and armour. The project produced a guidance toolkit which offers nuanced perspectives on the use of AI in museums. Awarded by the Collections Trust, its findings have been part of a UK parliamentary showcase and cited in an internal UNESCO report.
Asian and African Gallery, Royal Armouries, Leeds. © Royal Armouries
Of the two main factors in the project’s success, the first was being able to quickly mobilise networks of collaborators to both secure the funding and deliver the award. The project originated from speculative contacts in academia, made possible by the museum’s creation of a Research Manager role several years earlier. The academics, in turn, assembled a team with an impressive range of disciplines and specialisms, and drew in a wider network of collaborators who provided valuable insights.
Similarly, the museum was willing to be transparent about potentially embarrassing topics. This supported the project’s development of insights into the colonial nature of knowledge and the limitations of catalogue data for AI processing. Such openness was possible only by quickly building trust between the museum and its academic partners. Doing so created a valuable resource for museums navigating the complex landscape of AI.
XXVIA.87: Helmet, late 18th century; Hyderabad, India. Presented by the East India Company, 1851. © Royal Armouries