Cursor AI adoption delivers a significant, temporary surge in development velocity but simultaneously drives up static analysis warnings and code complexity, creating a long-term slowdown according to a new study. Researchers used a rigorous difference-in-differences approach comparing Cursor-using GitHub projects against matched non-Cursor projects, finding the velocity increase is statistically significant yet transient. Static analysis warnings and code complexity rise substantially and persistently, becoming the primary drivers of the subsequent velocity decline. This highlights a critical quality assurance bottleneck for early adopters of agentic AI coding tools like Cursor.