Research Group Meeting · August 5, 2026
HCII Montreal Conference Recap & Paper Discussion
A conference recap and research discussion spanning AI safety, explainability, ethics, privacy, and agent-based modelling.
Human–Computer InteractionAI SafetyAgent-Based Modelling
Overview
The session reflected on themes from the Human–Computer Interaction International conference in Montreal, including AI alignment and control, privacy, trust, and responsible system design.
It also introduced a paper using agent-based modelling to study AI transparency, an approach that is less common in HCI than survey, experimental, and interface-focused methods.
Key Themes
- Explainability, alignment, and control as central AI-safety concerns.
- Ethical and privacy questions in deployed human–AI systems.
- Agent-based modelling as a bridge between computational mechanisms and human behaviour.
- The practical value of conferences for comparing methods and identifying collaborations.
Main Takeaways
- AI safety and responsible AI have become prominent themes across HCI research communities.
- Conference discussions can reveal how the same technology is evaluated differently across disciplines and user groups.
- Agent-based models are promising for HCI when their assumptions and behavioural mechanisms are explicit.