Promoting the development of new quality productive forces is a key task and inherent requirement for China to realize high-quality economic development in the new stage of development. Enterprises are important carriers and commitment subjects for high-quality economic and social development. How to promote the development of new quality productive forces in enterprises has important theoretical and practical value. Based on the realistic background of accelerating the construction of intelligent manufacturing development ecology in China, this paper takes the national intelligent manufacturing demonstration project as a quasi-natural experiment, adopts text analysis method to construct the measurement index of enterprise new quality productivity, and then applies the dual machine learning method to explore the influence effect and mechanism of intelligent manufacturing strategy on the development of enterprise new quality productivity. It is found that intelligent manufacturing strategy can significantly promote the improvement of new quality productivity of enterprises, and this conclusion is still valid after a series of robustness tests including endogenous processing. Intelligent manufacturing is not equivalent to the new quality productivity of all kinds of enterprises. In non-state-owned enterprises, technology-intensive enterprises, areas with high intellectual property protection level and areas with perfect information infrastructure, intelligent manufacturing has a more obvious enabling effect on their new quality productivity. Intelligent manufacturing strategy mainly promotes the development of new quality productivity of enterprises through three channels: optimizing the human capital structure of enterprises, reducing the cost of information acquisition of enterprises, and easing the capital constraints of enterprises. From the perspective of the research, this paper explores the influence effect and action path of intelligent manufacturing on the level of new quality productivity of enterprises, which helps to enrich the relevant research on the development of new quality productivity, and provides theoretical and empirical evidence of China’s scenario for exploring the path of improving new quality productivity of enterprises from the perspective of industrial intelligence. In terms of research content, the effects of heterogeneity such as the type of enterprise ownership, the attributes of production factors, the level of regional property rights protection and the level of information infrastructure on the development of new quality productivity of enterprises are investigated, and more detailed empirical conclusions are provided for targeted policy recommendations. In terms of research methods, the use of the more advanced dual machine learning method can better avoid the bias and endogeneity problems of the traditional econometric model applied to policy causal inference. In terms of variable measurement, the text analysis method that comprehensively considers the word frequency of core terms connoted in new-quality productivity, such as new production technology, emerging production factors, advanced organization configuration, is adopted to construct the measurement index of enterprise new quality productivity, which overcomes the limitation that the existing index system method has too many index elements and easily fails to meet the axiomatic criteria such as unity, consistency and additivity. The research content and conclusion have reference value for China to seize the opportunity of intelligent manufacturing to enable enterprises to improve new quality productivity.