After deploying a persistent, proactive AI 'teammate' across multiple teams at a large technology company, researchers observed human-agent workplace boundaries being continually negotiated across collaboration rules, the non-human actor's relational boundaries, and the redistribution of trust.

The study identifies that the boundaries of the human-agent workplace are actively in flux, triggering breakdowns and negotiations across three areas: tacit rules of collaborative human workflows, the relational boundaries of this new non-human actor, and the redistribution of trust and human agency. Whereas enterprise AI has largely been single-user, reactive tools, this work turns to proactive, multi-user 'teammates' and provides an empirical account of the boundary negotiations this new form triggers in real teams. Based on an in-situ qualitative study of a persistent, proactive AI agent 'teammate' deployed across multiple teams in a large technology company, using early micro-negotiations as analytical signals.

The authors use these early micro-negotiations as signals to chart a new research, design, and organizational agenda aimed at intentionally preserving human agency in a workplace shared with non-human organizational actors. Moves from describing observed boundary negotiations to proposing an action agenda for research, design, and organizational practice centered on human agency. The agenda is derived from breakdowns and negotiations observed in the in-situ qualitative study and is framed as an early-signal, agenda-setting proposal.

The conclusions are oriented to enterprise settings in which a persistent, proactive AI agent 'teammate' is deployed across multiple teams in a large technology company, and are relevant to researchers, designers, and organizational managers concerned with human-agent collaboration rules, the role of a non-human actor, and trust distribution. The proposed agenda offers direction for subsequent research, design, and organizational practice on how to intentionally preserve human agency in a shared workplace.

The paper uses early micro-negotiations as signals and is an exploratory qualitative study; its findings characterize boundary flux and negotiation processes rather than stable conclusions. Readers may watch how these negotiations evolve across different teams and organizational cultures, how the redistribution of trust and agency stabilizes or changes over time, and how the proposed agenda translates into concrete design and management practice. The abstract does not provide the number of teams, deployment duration, or details of negotiation cases, which remain open questions.