Formal Framework for Aligning Agentic AI Power and Responsibility

Proposes a process-oriented modeling framework to prevent privilege escalation and unsatisfiable expectations in multi-agent systems. It analyzes the symmetry between operational processes and institutional states to ensure agentic responsibilities match their actual capabilities.