Agency is not autonomy: Responsibility under uncertainty in accelerated human–AI systems

This record contains a position paper examining human agency and responsibility in accelerated human–AI systems. The paper argues that contemporary governance approaches relying on ex-post accountability are structurally insufficient under conditions of speed and uncertainty. Instead, it proposes ex-ante governance through the deliberate design of pauses, procedural friction, and responsibility-preserving structures at decision time. The work emerges from ongoing conceptual and field-based research conducted within the Dom Ciszy research programme and is intended for research, policy, and interdisciplinary discussion.