Global AI Governance Radars and Agent-Level Public Evidence: A Comparative Observatory Note | Synapse