A half-day workshop at HAI 2026 November 16, 2026 (Monday), Morning
Ritsumeikan University, Osaka, Japan
We sincerely thank our speakers, Minseo Kim (Pusan National University) and Shoichi Nakanishi (Plus Circulation Japan), and everyone who joined us and took part in the discussion.
We look forward to continuing the conversation on agent facilitation!
The rapid maturation of large language models and generative AI has shifted artificial agents from passive tools toward proactive partners. This workshop takes up this shift under the theme "From Interaction to Agency."
HAI and adjacent communities have made real progress on human–agent mediation, collaboration, and teaming, in which an agent contributes to a shared task alongside people.
Far less attention has gone to agents whose role is not to contribute task content at all, but to shape how a group of humans interacts mediating, structuring, and unblocking the collective process itself.
We call this role Agent Facilitation, and we distinguish it from neighboring areas along a single axis: whether the agent's intervention targets task content or interaction process.
A facilitation agent stays content-neutral with respect to what the group decides, and acts instead on the process by which they decide: surfacing dissent, protecting psychological safety, redistributing turn-taking, interrupting premature convergence.
Designing Agent Facilitation (DAF) invites researchers from HAI, HCI, AI, cognitive science, psychology, sociology, management, and design research to map this space together.
We invite short papers of 1 to 4 pages (excluding references), written in English in the ACM SIGCHI format.
We especially welcome submissions reporting failure cases, limitations, and negative results — the kind of material usually filtered out of main-track papers, and precisely what a new sub-area needs in order to mature.
Work-in-progress and position papers are equally welcome.
Topics include, but are not limited to:
Agents that maintain psychological safety through non-verbal behavior
Agents that intentionally disrupt conformity pressure or group-think
Facilitation in AI-coexistence societies
Evaluation methodologies for group-facing (not user-facing) agents
Ethics, autonomy preservation, and the politics of agent-mediated decisions
Speculative- and SF-prototyping-informed design of facilitation agents
Failure cases and limitations of agent facilitation in field deployments
Important Note on Publication:
Papers accepted to this workshop will NOT be published in the HAI 2026 ACM proceedings.
Submission Deadline: October 29, October 19, 2026 (23:59 AoE).
Please apply to be a presenter at this workshop here: Google Form Link
Notification of Acceptance: October 31, October 26, 2026 (23:59 AoE).
Important Note on Publication:
Papers accepted to this workshop will NOT be published in the HAI 2026 ACM proceedings.
Tomoya Minegishi (Tokyo Online University)
minegishi.tomoya [ a t ] internet.ac.jp
Toshio Mochizuki (Waseda University)