ImAIEthics will investigate the role visual representations of Artificial Intelligence (AI) have in the
constitution of the technological imaginaries about AI. Technological imaginaries are not detached from the
technological reality of AI; they play an active role in the innovation and implementation of AI-based systems
in our societies, insofar as they frame the expectations of engineers, citizens, and stakeholders. The project
will focus on the ethical implications of the use of images of AI in institutional science communication.
Indeed, there is a gap between the current debate about AI ethics and the lack of attention to the ethical
consequences of the way AI is presented to the public at large. Moreover, specific attention will be devoted
to the abundant use of stock images, i.e., pre-produced images made available for license on the websites of
agencies such as Getty Images and Shutterstock. These images represent AI in terms of white robots, halfflesh, and half-circuit blue brains, etc. It is commonly believed that stock images do not deserve our attention,
because they are just the “wallpaper” of the consumer culture. However, stock images are today pervasive:
stock images of AI are widely used in official contexts such as institutional websites, reports of Europe-funded
research, etc. ImAIEthics will ask questions such as: what kind of expectations and sentiments do these images
provoke? How the use of these images relates to principles of AI ethics like transparency and trustworthiness?
Is there a better way (from an ontological, aesthetical, and ethical point of view) to visually represent AI? Is
there an intrinsic problem with the visual representability of AI? ImAIEthics will bring together methods and
ideas from different research fields. Moreover, the project will combine theoretical and ethical considerations
with the empirical analysis of two case studies: the catalogs of Getty Images and Shutterstock, and the
websites of the European database CORDIS and the US National Research Foundation