Some in tech had predicted that the rise of single-person AI-enabled unicorns, and while that hasn’t quite yet materialized, there seems to be some sort of a thrust in the general direction.
A new working paper titled “AI as “Co-founder”: GenAI for Entrepreneurship” supports this sentiment, presenting evidence that the diffusion of Generative Artificial Intelligence is fundamentally reshaping the entrepreneurial landscape. The research posits that GenAI is effectively acting as a digital co-founder, automating core tasks and lowering the entry barriers for new ventures. To establish a causal link between GenAI diffusion and firm creation, the researchers employed a difference-in-differences framework, exploiting the unexpected November 2022 release of ChatGPT as a singular, global technological shock that sharply reduced start-up costs. Their methodology involved analyzing universal data on over 12 million firm registrations across China from 2021 through 2024, comparing changes in firm entry between hyper-local geo-coded grid cells—at a spatial resolution of roughly 5 square kilometers—that either had high or low pre-existing AI-specific human capital.
The study’s results demonstrated a striking asymmetry in new firm growth by size, validating the hypothesis that AI enables leaner ventures. Geographic grids with stronger AI-specific human capital experienced a pronounced rise in new firm entry following the launch of ChatGPT, a trend driven entirely by small firms. Conversely, the entry rate of large firms actually declined in these same high-AI areas, consistent with a structural shift toward organizational models that require less upfront capital. Further mechanisms explored in the paper revealed that these new firms were created with smaller amounts of capital, fewer shareholders, and smaller founding teams. The effect was strongest among first-time entrepreneurs, indicating that GenAI substitutes for the experience and managerial know-how typically possessed by serial founders. The growth was concentrated in AI-downstream sectors, such as retail, business services, online platforms, and digital-service industries, where GenAI tools can be readily applied for information-processing and customer-facing activities.
These findings carry significant implications for the future of business and technology. By demonstrating that GenAI relaxes critical constraints like the need for prior entrepreneurial experience, external financing, and managerial labor, the technology appears to be democratizing entrepreneurship. GenAI automates tasks such as coding, content creation, and legal document drafting, thus reducing the initial need for large teams and capital. This allows small teams to replicate capabilities that once required larger organizations, suggesting a future where business growth may be fueled by a larger volume of nimble, resource-constrained start-ups. For the tech sector, the research underscores that the most immediate entrepreneurial impact is felt in application-focused industries rather than in capital-intensive upstream sectors. Mark Zuckerberg has said that AI will give everyone the power of a 10,000 person company, and Anthropic’s Dario Amodei had said that a single-person unicorn will be a reality by 2026. And if these trends hold, a single-person unicorn — with an AI cofounder — might actually end up being created over the next twelve months.