We have identified two major strands of A.I. under development (and roll-out) now. Commercial/Economic A.I. and Strategic A.I. The differences between them are real, and complicate their governance. A governance regime must take these differences into consideration. Here is a brief comparison between the two.

Dimension	Commercial / Economic AI	Strategic AI
Primary objective	Productivity, revenue, efficiency, innovation	National security, military advantage, strategic influence
Main actors	Corporations, investors, consumers, regulators	States, militaries, intelligence agencies, defense contractors
Principal risk	Economic and social externalities	Security escalation and strategic instability
Governance problem	How to make markets innovative but accountable	How to make powerful systems controllable without creating strategic vulnerability
Typical failure	AI causes discrimination, market concentration, job disruption, fraud	AI makes or accelerates consequential decisions involving conflict, deterrence, intelligence, or weapons
Transparency	Often desirable	Often constrained by national security
Competition	Generally encouraged	Can become a national-security imperative
International coordination	Trade, standards, privacy, competition law	Arms control, strategic stability, international law
Accountability	Regulators, courts, shareholders, consumers	Civilian governments, military command structures, international law