This article applies survival analysis and network methods to study how 966 generative artificial intelligence (AI) startups have progressed through funding stages, drawing on Crunchbase data from 2015 through mid-2025. The headline finding is a 21.4% seed-to-Series-A progression rate, which sits comfortably within the range observed in other technology sectors once the post-2022 funding environment is considered. This conformity to established patterns is itself the central contribution: despite unprecedented hype and capital inflows, generative AI venture funding dynamics do not appear fundamentally different from those of prior technology waves. The methodological contributions are secondary but potentially useful. Cox proportional hazards modeling is employed to identify predictors of faster progression, and a knowledge graph of 1,924 investors linked by 5,060 co-investment relationships is constructed to map how deals flow through the ecosystem. A predictive model using only seed-stage characteristics achieves 0.82 AUC (area under the curve), with accelerator backing and strategic investor involvement contributing most substantially to predictive power. Sensitivity analysis on survivorship bias suggests true progression rates could be 3%–9% lower than observed values. The broader implication is that investors and entrepreneurs can draw on existing frameworks for evaluating generative AI opportunities, rather than requiring entirely new analytical approaches for this emerging sector.
Jitesh Gurav (Mon,) studied this question.