A workshop on contribution, originality, and accountability in agent-mediated collaboration
LLM agents are now embedded in how teams research, design, write, and decide. They operate inside the moment when contributions are being formed. This workshop confronts what we call contribution dissolution: the blurring of attribution, originality, and accountability in agent-mediated collaborative work.
Collaborative knowledge work rests on an implicit social contract: people who work together maintain a shared, informal/formal record of who contributed what. We call this witnessed contribution — work that can be attributed, contested, and accounted for because it emerged in a social situation that made it intelligible to others.
Witnessed contribution is not the same as documented provenance. A provenance log records that something was produced and how. By itself, however, such documentation cannot ensure accountability, since accountability also depends on contributors making recognizable judgments that others can interpret, question, and evaluate.
LLM agents shift this. People now co-think with agents, developing reasoning in dialogue rather than externalizing already-formed ideas. The boundary between what the individual contributed and what the agent introduced is never drawn. Documentation cannot recover what the process never produced.
When agent-mediated work circulates between collaborators, neither sender nor recipient can reliably distinguish human contribution from agent reformulation. The shared understanding that makes collaboration productive begins to blur and break down.
How do agents restructure the process through which contributions are formed, and what is lost when they intervene inside the collaborative loop? Both at the individual level (uncertainty about what is genuinely yours) and the collaborative level (collapse of productive friction between distinct perspectives).
\Why do documentation-based responses, such as AI use statements, detection tools, watermarking, and provenance tracking, fall short, and what does this reveal? They assume that the evidence needed for attribution already exists and merely needs to be surfaced. But what if that evidence was never produced?
If documentation cannot recover what was lost, what would accountability infrastructure require? What collaborative systems, governance mechanisms, and institutional norms would need to preserve for witnessed contribution to remain possible in agent-mediated work?
We invite 2–4 page position provocations from researchers, practitioners, and theorists working on human–AI collaboration, accountability in cooperative work, or the social organization of knowledge production. Two deadlines so participants who need lead time for visas or travel can plan ahead.
For participants who need advance planning for travel, visa applications, or institutional approval. Reviews begin as submissions arrive.
Due to continued interest from the community, we have extended the submission deadline to allow additional time for prospective participants to prepare their submissions.
| Date | Milestone |
|---|---|
| 2026-07-10 | Early submission deadline (AoE) — [rolling review] |
| 2026-08-11 | Extended submission deadline (AoE) |
| 2026-08-14 | Final acceptance notifications |
| Sunday, October 11, 2026 | Workshop day — CSCW 2026 |
ALL DEADLINES ANYWHERE-ON-EARTH (AoE). REGISTRATION FOR CSCW 2026 IS REQUIRED TO ATTEND.