OPEN-WEIGHT AI MODELS WON'T REPLACE CLOSED RIVALS

The AI infrastructure buildout faces persistent constraints that will shape the competitive landscape for years to come. Labor shortages and power limitations remain critical bottlenecks in expanding data center capacity, preventing a glut of supply that might otherwise favor cheaper open-weight alternatives. Corporate adoption of AI continues to generate measurable returns, fueling sustained demand across sectors. This sustained appetite means multiple model types will remain viable—enterprises will likely maintain portfolios combining open-weight models for cost efficiency and control with closed commercial models for cutting-edge performance and support. The Morgan Stanley analysis suggests the competitive dynamic won't collapse into a winner-takes-all scenario. Instead, infrastructure constraints and diverse use cases will support a durable ecosystem where open-weight and closed models coexist, each serving different organizational needs and budgets.