Tackling Role Transitions in Work with Intelligent Personal Assistants

The integration of agentic artificial intelligence into information systems has created a demand for humans to play AI-facing roles. We report on a study that examined interaction with intelligent personal assistants (IPAs), AI-based systems that give feedback and guidance on activities such as time management, wellbeing, and communication. The study investigated users’ emergent roles as AI partners, challenges involved, and capabilities required to benefit from interacting with IPAs in practice. Drawing on role theory and boundary-spanning theory, we find that when workers interacted with IPAs, their workflow fractured into sequences of micro-tasks, requiring them to undertake frequent micro-role transitions between human-facing and AI-facing roles. This suggests that boundary-spanning competence - the ability to shift fluidly between human-facing and AI-facing roles and logics – and rapid reflexive practice are key capabilities for workers to integrate AI-facing roles into work and create value through hybrid intelligence.