The AI Ecosystem Dissected: A Comparative Empirical Analysis of AI Firm Business Metric Scaling

Abstract

Artificial intelligence (AI) has after numerous cycles of hype and disillusionment experienced a surge in relevance and popularity, fueled by readily available, advanced deep learning models. The private sector has taken to it with great interest, spurring both adoption and innovation. However, firms involved in AI—whether adopting, enabling, or producing—differ in important ways along business models, operations, and scaling potentials. This nuance is often overlooked categorization attempts. This paper argues that a comprehensive understanding of the AI ecosystem, including its digital capabilities and distinct actors, is critical for unraveling disparate value capture mechanisms. The study further explores sectoral and regional dynamics, examining their unique opportunities and barriers for AI enabled firms as a whole, and at a more granular level. An empirical sample and analysis of AI-enabled firms was conducted, with business metric scaling as a proxy for firm financial outcome. Findings reveal that regional positioning significantly impacts only the largest players, while sectoral effects on business metrics for AI-enabled firms vary widely, not always favoring more digitized sectors. Focusing on individual categories, scaling variations increase, particularly for AI laboratories, which are heavily influenced by sector choice. The paper also presents a novel AI-driven labeling approach, arguing its superior performance compared to standard industry NLP solutions.