AutomationXP26 - Workshop@CHI26

Recent advances in Artificial Intelligence (AI) have enabled agentic AI systems that coordinate multiple specialized agents behind unified interfaces. These systems can independently initiate actions and solve complex problems. In traditional automation systems within organizations, workers maintained clear oversight - they could see which system handled each task and trace outcomes to specific processes. The integration of agentic AI, however, obscures this relationship and makes it more difficult for humans to identify which agent is responsible for a given outcome. This creates novel research challenges in the field of “Automation Experience”, particularly in terms of transparency, human agency, and long-term human-AI collaboration dynamics

The objectives of the workshop are as follows:

Examine challenges and opportunities in interaction with multi-agent systems,
Develop theoretical frameworks for understanding agency and temporal evolutions in organizational agentic AI systems,
Identify promising future research topics in the form of project ideas and a research agenda,
Expand and strengthen a multi-disciplinary network of automation experience researchers.

Recent and Previous Activities of the Automation Experience Initiative

In the workshop, we will elaborate on the following three core challenges regarding agentic automation experiences. Example research questions include:

Enabling Multi-Agent Transparency and Attribution

What attribution mechanisms can help humans understand which agents contribute to specific outcomes in multi-agent systems?
How much transparency into multi-agent coordination do humans need for effective collaboration, and when does additional transparency become counterproductive?
How can interfaces be designed to provide transparency into multi-agent coordination without overwhelming humans?

Balancing Human Agency in Agentic AI

How do humans maintain meaningful agency when multiple agents coordinate to solve complex tasks with minimal human input?
What level of human involvement preserves appropriate agency while enabling effective multi-agent problemsolving?
What interface designs support human agency when multi-agent coordination handles complex organizational tasks?

Sustaining Human Skills in Long-Term Human-AI Interaction

How do human skills and expertise change over time when agentic AI handles complex problem-solving with minimal human involvement
How can organizations maintain human competence while enabling long-term agentic AI collaboration?
What intervention strategies prevent deskilling while preserving the benefits of agentic AI autonomy?

Agentic AI has rapidly gained traction and are increasingly integrated across organizations: they shift automation experiences from specific tools to proactive systems coordinated by multiple specialized agents, which are able to solve complex problems with minimal human involvement. This workshop aims to examine the critical challenges posed by agentic AI in organizational settings, particularly regarding enabling multi-agent transparency and attribution, balancing human agency in agentic AI, and sustaining human skills in long-term human-AI interaction. We invite position papers based on recent or forthcoming research, case studies, or design work. We encourage participants to share insights from real-world implementations and challenges of multi-agent coordination, agency, and evolving human-AI relationships.

Topics of interest include but are not limited to

Attribution mechanisms for multi-agent coordination
Balancing human agency with autonomous multi-agent problem-solving
Understanding when reduced agency improves versus harms human well-being and outcomes
Maintaining human expertise and preventing deskilling in long-term agentic AI collaboration
Real-world organizational cases of agentic AI adoption
Socio-technical considerations for integrating agentic AI into workflows

Submission and Participation

Submissions should follow the single-column ACM conference proceedings format and may not exceed five pages (excluding references).
Non-anonymized papers must be submitted in PDF format to https://easychair.org/conferences/?conf=automationxp26.
The organizing committee will evaluate submissions based on their relevance, originality, significance, and quality.
Upon acceptance, at least one author of each accepted position paper is required to attend the workshop.

Important Dates

Submission of position papers: February 20th, 2026 (extended!)
Decision to authors: February 27th, 2026
Camera-ready versions due: March 27th, 2026
Workshop: April 14th, 2026

Short Bio

Zana Buçinca is a Postdoctoral Researcher at Microsoft, and an incoming Assistant Professor at MIT, with a shared appointment in Sloan and EECS. She earned her PhD in Computer Science at Harvard, where her research at the intersection of human-AI interaction and responsible AI integrated cognitive and social science theories to design novel interaction techniques that complement workers and amplify their values in AI-assisted tasks. Her work has been recognized with the IBM PhD Fellowship, a Siebel Scholarship, and Best Paper Awards at CHI and IUI. She has also been named a Rising Star in AI by the University of Michigan, a Rising Star in Management Science & Engineering by Stanford, and one of the Top 10 Most Inspiring Women in STEM by UNDP Kosovo.

CEST
14:15 – 14:30	Welcome and introduction
14:30 – 14:55	Lightning Talk by Zana Buçinca
14:55 – 15:25	Paper Madness I
15:25 – 15:45	Discussion I
15:45 – 16:30	Break
16:30 – 17:00	Paper Madness II
17:00 – 17:20	Discussion II
17:20 – 18:00	Synthesis and closing