LEXFLOW 2027 invites contributions on intelligent interfaces for human–AI legal reasoning. Taking place alongside ACM IUI 2027 in Helsinki, the workshop will explore how people can understand, supervise and steer legal workflows involving multiple AI agents, tools and evidence sources.
LEXFLOW 2027: Workshop on Agentic Workflows for Human–AI Legal Reasoning will take place on 8 February 2027 in Helsinki, Finland, co-located with the 32nd International ACM Conference on Intelligent User Interfaces. The workshop welcomes researchers and practitioners interested in the design, evaluation and use of intelligent interfaces for legal AI.
As AI becomes embedded in legal research, drafting, compliance and fact verification, legal professionals increasingly encounter systems that operate across multiple stages. A single task may involve planning, document retrieval, source comparison, intermediate judgments and coordination among specialised agents. Understanding a final answer therefore requires attention to the process through which it was produced.
These developments create important questions for interface design. How can users identify the sources supporting a conclusion? Where should they intervene when an agent follows an unsuitable approach? How should uncertainty, conflicting evidence and differences between agent contributions be presented? What information enables a professional to review a workflow and take responsibility for its results?
LEXFLOW 2027 will examine interaction mechanisms that make these processes accessible to users. Topics include evidence provenance, source verification, uncertainty communication, workflow decomposition, multi-agent coordination, contestability and correction. The workshop will also consider evaluation methods that assess whether interfaces enable meaningful oversight in realistic legal tasks.
The programme will combine an invited talk, paper and demonstration presentations, a workflow gallery, a collaborative design sprint and structured discussion. Participants will share Workflow Cards describing legal tasks, human and AI roles, evidence sources, intervention points and potential failures. These activities will support the development of reusable design patterns and a shared research agenda for human-centred legal AI.
The organising team comprises Sirui Han, Yuyao Zhang, Yidan Huang, Chuxue Cao, Yujin Zhou and Guoying Lu from HKUST.
The workshop welcomes research papers of 10 or more pages, short papers of 5–9 pages and demo papers of 5–9 pages. Contributions may include empirical studies, interactive systems, conceptual frameworks, methodological work and early-stage design explorations.
Important dates
Submission deadline: 9 November 2026, 23:59 AoE
Notification: 7 December 2026
Camera-ready deadline: 21 December 2026
Workshop: 8 February 2027
Further information and submission instructions: https://lexflow2027.github.io/
https://lexflow2027.github.io/
All events and news