To content
Research

We are at BPM 2026

The Chair presents three contributions on agentic process modeling, addressing erroneous AI decisions, and AI-supported BPMN education

The BPM conference is one of the world’s leading events in Business Process Management and Process Mining. This year, it takes place from September 27 to October 2 at York University in Toronto, Canada. Our team is represented by three contributions that offer different perspectives on the use of artificial intelligence in process management. They focus on collaboration between humans and autonomous AI agents, the responsible design of autonomous processes, and support for learners. Further information is available on the conference website.

Our contributions at a glance:

Agentic Business Process Modeling: Co-Creating Business Processes with Human-Centric AI Agents
Seyyid A. Ciftci & Christian Janiesch

"How can autonomous AI agents and humans jointly develop process models without compromising human decision-making authority and critical engagement? The contribution develops and evaluates an agentic system that independently plans and executes modeling tasks and reviews its results. Collaboration is designed to enable humans to focus on reviewing the content and making targeted revisions. Visual feedback helps them identify inconsistencies and make informed decisions. The evaluation shows higher modeling quality and lower perceived workload compared with earlier AI configurations. These findings inform design principles that combine technical performance with human autonomy and reduced cognitive load."

You Shall Not Pass Unchecked: Redress of Rogue AI in Autonomous Processes
Christian Janiesch

"When AI agents independently make decisions and execute tasks, those affected must be able to challenge erroneous or unjust outcomes. The contribution distinguishes three mechanisms: objections, contestations, and formalized review procedures. Under the concept of “redress,” it considers broader remedies that go beyond merely reversing individual process steps. The article shows why such mechanisms must be considered from the outset when designing autonomous processes to ensure human agency, accountability, and fair outcomes."

BPM-Tutor: A Conversational, AI-based Learning Platform for BPMN Modeling Education
Seyyid A. Ciftci, Roman Geist & Christian Janiesch

"With the BPM-Tutor, we offer a conversational, AI-based learning platform for BPMN process modeling. Five configurable AI roles support learners in different ways: from targeted questions and explanations to the joint development of process models. Visual feedback directly on model elements helps learners understand errors and make targeted improvements to their models. We successfully used the platform in our “Business Process Management” course in 2026 to make learning BPMN easier for students. Additional features support task management, assessment, and the conduct of empirical studies. Interested users can explore the platform through a publicly accessible demo version."