The AI-2040 project outlines a strategy where the US and China cooperate to decelerate transformative AI development. This "Plan A" relies on layered defenses, including centralized compute in neutral countries under military surveillance. Practitioners should focus on the specific verification mechanisms and incentive structures required to maintain such a fragile geopolitical truce.