Collab AI Research
How Microsoft is using science and research to invent the future of collaboration
Welcome to Collab AI
Collaboration is a central feature of how work gets done, involving coordination across people, tasks, and information sources. As AI systems increasingly contribute to shared workflows—synthesizing discussions, generating alternatives, maintaining context, and supporting group decision processes that unfold across both synchronous and asynchronous interaction—understanding how humans and AI function together in group settings has become a scientific and operational priority. This shift raises new questions about evolving intent, shared grounding, and how collective reasoning unfolds across time and modalities.
Collab AI serves as Microsoft’s research center dedicated to studying collaborative processes and building intelligent systems capable of participating in them. The work spans empirical studies of team interaction, computational models of shared context, dialogic representations that capture how knowledge is co‑created, and methods for enabling AI to engage in multi‑party alignment and decision‑making. Featured papers, talks, and research outputs explore emerging data, architectures, and evaluation frameworks for systems that operate not as single‑user tools, but as contributors to group activity.
Dive in to explore the science behind collaborative AI—and to engage with research aimed at creating intelligent systems that do more than assist individuals. The goal is AI that helps teams form collective purpose, sustain coherent dialogue, and work together more effectively than ever before across conversations, shared and co-authored artifacts, meetings, and organizational workflows.
Evaluating Proactive AI Mediators in Multi-Party Conversation with ProMediate
As LLMs evolve toward handling multi-party teamwork, they are increasingly expected to navigate complex group dynamics. However, current research often overlooks the nuance of these interactions. While agents are being designed for multi-party settings, we still lack the benchmarks to evaluate the real-time effectiveness of their interventions or their broader socio-cognitive intelligence when working with groups of people.