Abstract
We set out to enquire how people who interact with a virtual advisor, an artificial intelligence-powered (AI) chatbot, at a large bank in Northern Europe experience their encounters with it. We contend that anthropomorphism, the tendency of people to humanise technologies such as AI, shapes the experience of the users in this context and subtly informs interactions between humans and the bot. In the organisation we studied, humans interacting with AI did so in silos without engaging in boundary work towards each other. Additionally, humans purposefully accommodated and gave symbolic space and meaning to the AI chatbot, whose presence was expanding within the organisation as they intermingled with it in their daily work. We suggest that boundary work involving humans in different roles and humanised AI technology is largely missing in many organisations that are investing in digital transformation to leverage its many promises.