Purpose: In today's rapidly changing digital world, the ability of startups to grow has become a major issue. This research looks at how AI-enhanced Human Decision-Making, AI-powered Entrepreneurial Agility and Ethical AI Governance help startups create an AI-driven Innovation Ecosystem that improves AI Infrastructure Readiness and supports both internal and external startup scalability potential.

Design/methodology/approach: The Prolific platform was used to collect data from 274 decision-makers who worked for AI-intensive firms in the US. We used modified Likert-scale items from previous validated research to measure constructs. Partial Least Squares Structural Equation Modelling in SmartPLS4 was used to test the model. Both direct and indirect effects were analysed to examine the hypothesized relationships.

Findings: All AI-enabled capabilities had a strong positive association with AI-driven Innovation Ecosystem. AI-driven Innovation Ecosystem significantly influenced AI Infrastructure Readiness, which in turn positively impacted both Startup Scalability Potential – Internal and Startup Scalability Potential – External. Mediation investigations showed that AI-driven Innovation Ecosystems and AI-Infrastructure Readiness serially mediate the impact of AI-enabled capabilities on startup scalability.

Practical implications: Theoretically, this study builds on the Dynamic Capabilities Theory by adding AI-driven ecosystem and infrastructure preparedness as sequential mediators. This gives us a new way to look at how businesses use AI to develop quickly. In practice, the results give startup leaders and policymakers useful information on how to build AI adoption plans that are not only new but also follow the rules and fit with the infrastructure.

Originality/value: This article presents a new serial mediation paradigm that connects AI capabilities to startup scalability and backs it up with real-world evidence from the US, a nation that is driven by innovation.