US Strategy for General Artificial Intelligence Competition: Insights and Implications

On February 23, 2026, the RAND Corporation released a research report titled “Decisive Economic Advantage: Modeling the Transition from Temporary First-Mover Leads to Economic Dominance in Artificial General Intelligence.” The report focuses on the geopolitical competition landscape in the global AGI field and introduces the Decisive Economic Advantage (DEA) analytical framework. Through dynamic economic models and thousands of Monte Carlo simulations, it systematically explores the core conditions, driving mechanisms, and policy intervention effects for the transformation of AGI from a temporary first-mover advantage (FMA) to economic dominance, while analyzing the deep impacts of related competitive dynamics on the global technology landscape and economic security.

Core Analysis Framework and Research Design of US AGI Competition Strategy

Core Concept Definition

The report’s core innovation is the introduction of the Decisive Economic Advantage framework, defined as a competitive state where the capabilities of AGI, actual industry deployment, and capital reinvestment create a positive feedback loop. This leads to an ever-widening gap between leaders and followers, ultimately preventing followers from restoring competitive balance through policy adjustments or market competition during the relevant strategic cycle. This concept focuses on the endogenous dynamics of the economic system rather than a single technological threshold breakthrough, distinguishing itself from the traditional military-centric framework of “decisive strategic advantage” by emphasizing long-term economic competition.

The report identifies three core competitive dimensions: 1) Technology Quality Gap: the productivity levels of AGI systems among different entities; 2) Deployment Intensity Gap: the breadth and depth of AGI integration into the economic production system; 3) Hardware Capacity Gap: the scale of physical infrastructure such as computing chips and data centers that support AGI operations. The dynamic interaction of these three dimensions directly determines the evolution direction of income disparities among competitors. The report also clarifies two core differentiation mechanisms: the feedback loop where expanded deployment generates application data that enhances technological capabilities, leading to broader deployment; and the hardware-reinvestment moat, where economic returns convert into hardware capacity expansion, further driving technological progress through deployment.

Competitive Dynamics Model and Four Analysis Scenarios

The report constructs a dual-block dynamic economic model, categorizing competitors into leading and following economies. The core assumption is that sufficiently advanced AGI can fully replace human labor in production, aligning with mainstream definitions of AGI’s ability to “perform all economic value tasks of humans.” The report builds a 2×2 basic analysis framework based on capital adjustment speed and hardware constraints, delineating four differentiated competitive scenarios that clarify the logic of competition differentiation under different institutional environments and hardware conditions.

Core Findings of AGI Competition Simulation

Through simulation analysis, the report identifies four typical competitive trajectories, clearly presenting the evolution patterns of disparities under different mechanisms: 1) Scale Barrier Type: technological, deployment, and hardware advantages rapidly diminish, leading to a quick convergence of leaders and followers to competitive equilibrium, completely dissipating first-mover advantages; 2) Intelligence Explosion Type: self-reinforcing technology accelerates the iteration of the leader’s AGI system, exponentially widening the gap across all dimensions; 3) Development Flywheel Type: while lacking self-reinforcing technological features, strong cross-feedback between deployment and technological learning still results in significant income differentiation; 4) Stable Gap Type: the gap between leaders and followers persists over time without infinite expansion, maintaining a limited asymmetric competitive state.

Overall simulation results show a strong regularity in the probability distribution of competitive convergence and differentiation. The duration of first-mover advantages primarily affects the severity of differentiation rather than whether differentiation occurs. The report concludes that convergence trajectories account for about 20% across all scenarios, with the vast majority of parameter combinations indicating a trend of differentiation in AGI competition rather than natural convergence. Rapid capital adjustments accelerate the arrival of extreme differentiation, while slow adjustments only delay the speed of gap expansion without altering the final trend of differentiation.

Core Driving Mechanisms of Decisive Economic Advantage

The report dissects all extreme differentiation trajectories with a 100-fold income gap, identifying three core driving mechanisms and clarifying the occurrence probabilities and core characteristics of different mechanisms. Results indicate that technological intelligence explosion is the primary source of long-term economic dominance in the AGI field, while investment and accumulation in hardware infrastructure are the core paths for achieving competitive differentiation in non-self-reinforcing scenarios. Additionally, the report notes that development flywheels are more of a starting point for differentiation rather than a long-term dominant mechanism, as most differentiation driven by development flywheels ultimately shifts towards the hardware-reinvestment cycle.

Effects and Boundaries of Policy Interventions

The report designs two types of policy intervention prototypes: “full-stack blockade” and “ecosystem containment.” The former targets restrictions across the entire chain of technology, hardware, and investment, while the latter focuses on limiting third-party market deployments and weakening development flywheel effects, leading to three core conclusions:

1. Timing of intervention is far more important than intervention strength. Implementing interventions before income disparities reach five times is much more effective than acting after the gap widens. As economic asymmetries grow, the marginal benefits of strategic interventions sharply decline, and waiting for definitive evidence can cause the loss of effective intervention windows.
2. Intervention effects are highly dependent on the driving mechanisms of differentiation. Differentiation driven by intelligence explosions is highly resistant to policy interventions; even with strong full-stack blockades, only 11.5% of cases can avoid extreme differentiation. In contrast, differentiation driven by accumulation (hardware-reinvestment cycles, development flywheels) is more sensitive to policy interventions, with moderate full-stack blockades able to prevent extreme differentiation in 67.5% of cases, and strong interventions achieving a 92.8% avoidance rate.
3. Different policy tools exhibit significant differences in effectiveness. The effects of full-stack blockade policies are significantly superior across all scenarios compared to ecosystem containment policies, which only produce weak effects in accumulation-driven differentiation and have almost no impact on intelligence explosion-driven differentiation.

Dual Impact of AGI Competition Landscape and Global Insights

Dual Impact on Global Technology and Economic Landscape

From a positive perspective, the analytical framework provided by the report offers a systematic view for countries to understand the competitive dynamics of AGI. The identified critical conditions for “convergence-differentiation” can guide countries in formulating differentiated AGI development strategies, while also promoting global attention to key issues such as AGI technology diffusion and infrastructure development, providing theoretical support for global technology governance.

From a risk perspective, the mechanisms for forming decisive economic advantages may exacerbate a zero-sum game tendency in the global AGI field. Leading economies might adopt more aggressive technology blockades and export control measures based on the judgment of solidifying first-mover advantages, undermining the stability and collaboration of global semiconductor and AI supply chains. Additionally, the potential for extreme differentiation could intensify technological arms races among countries, weakening international cooperation in the AGI field and further expanding the technological gap between developed and developing countries, creating new digital economic hegemony.

Moreover, the report’s model confirms that excessive technological controls and blockades can ultimately backfire on leading economies. Obstructing technology diffusion weakens the market scale and revenue capabilities of leading firms, reducing their ability to invest in R&D, while artificially severing supply chains increases operational costs for companies, ultimately inhibiting the innovation speed of AGI technologies in the long run.

Insights for Global AGI Governance and National Strategies

1. Competitive monitoring should focus on institutional transformation rather than single technical indicators. For countries, the core judgment criteria for AGI strategies should not be limited to single technical milestones like model parameters or computing power scale, but should also pay attention to signals of competitive system transitions from convergence to differentiation, monitoring leading indicators such as deployment scaling speed, hardware investment intensity, and deployment-technology transformation learning effects to anticipate changes in competitive landscapes.
2. Strategic design must match differentiation driving mechanisms, avoiding one-size-fits-all policy layouts. Different strategies should be developed for the two distinct paths of technological breakthroughs and accumulation-driven differentiation. For frontier technology competition, increased investment in basic research and improved collaboration between academia and industry are necessary; for the hardware-reinvestment accumulation path, strengthening computing infrastructure and optimizing supply chain layouts while expanding AGI’s industrial application scale to create a positive cycle of “application-technology-investment” is essential.
3. Build rapid response capabilities under uncertainty to seize strategic intervention windows. The report confirms that effective intervention windows are highly limited, and the value of early responses far exceeds that of late strong interventions. Countries need to enhance the decision-making infrastructure for AGI strategies in advance, pre-setting policy options and response plans to improve the flexibility and responsiveness of strategic decisions, taking timely action at critical junctures to avoid falling into a prolonged competitive passive state.
4. Promote multilateral collaboration in global technology governance to avoid vicious competition and landscape differentiation. As a transformative technology with global implications, the development and governance of AGI require collaborative efforts from the international community. Countries should promote the establishment of multilateral dialogue mechanisms and cooperate in areas such as technology diffusion, standard-setting, and risk prevention, curbing tendencies toward technology blockades and zero-sum games, while also narrowing technological gaps between nations through technology assistance and capacity building, ensuring that AGI technologies benefit the global community.