Purpose While artificial intelligence (AI) is a hot topic in public debate and academic discourse, most research is still conceptual with only a few insights into how AI is used in entrepreneurship practice. This study explores empirically how and with what effect early-stage IT startups utilize AI tools, with the general aim of getting a better understanding of how a new technology is adopted in the entrepreneurial process. Design/methodology/approach Conceptually, our study builds on the External Enabler Framework to structure the analysis. Following a mixed-method approach, we first analyze qualitative data on how AI tools are used in early-stage startups and identify the mechanisms that are facilitated. In a quantitative study, we test the effect of two identified efficiency mechanisms. Specifically, using Crunchbase data, we compare startups before and after the introduction of ChatGPT-3, focusing on (1) the number of employees and (2) the time to achieve seed funding. Findings Our qualitative study indicates that startups use AI mainly in the form of GenAI tools to streamline the process of venture creation, saving time and resources, and less frequently for directly shaping the offered product or the venture itself, presumably resulting from the distinct demands of different applications. While AI is considered important, it has so far not replaced human agency. Building on these results, the quantitative study confirms that startups with access to GenAI tools require fewer employees and achieve critical milestones faster, specifically securing seed funding. Research limitations/implications Our qualitative study is based on interviews with early-stage IT startups applying AI tools. In the quantitative study, it is challenging to distinguish between AI-infused efficiency effects and investor-related supply-side effects, resulting from the hype surrounding AI. Future research is needed to explore AI's broader influence on entrepreneurial processes. Originality/value Our study is one of the first to uncover the use of AI tools in early-stage startups, allowing entrepreneurs to compare with others and enabling policymakers to identify changes in the startup process, with implications for policy design. For academics, our study contributes to the discussion around AI and agency in the entrepreneurial process and provides insights into the applicability of the External Enabler framework.
Bergmann et al. (Thu,) studied this question.