Opportunity in the Age of Agents: How Generative AI Changes Opportunity Recognition, Evaluation and Creation

Entrepreneurship theory explains how opportunities are recognised, evaluated and created, but it models every stage of that process as exclusively human. Generative artificial intelligence (AI) now participates in all three stages: founders use large language models as co-ideators that generate opportunity candidates and as drafting partners that develop them, while investors use the same class of models to screen founders' submissions. This conceptual paper integrates opportunity theory (discovery, creation and effectuation) with the emerging empirical evidence on generative AI and ideation, creativity, evaluation and entrepreneurial performance. It argues that the co-ideator changes the opportunity process in four ways: it expands each founder's supply of candidates while compressing ideation skill gaps; it correlates opportunity pools across founders who draw on shared models, creating a homogenisation risk at market level; it raises the returns to prior knowledge, which governs selection among abundant machine-generated candidates; and it migrates durable advantage from generating opportunities to selecting, combining and enacting them. These claims are stated as seven formal propositions with boundary conditions, including the "jagged opportunity frontier" and the evaluation asymmetries produced by narrative AI explanations, with implications for founders, investors and educators, and a research agenda.