In recent years, corporate innovation activities have increasingly exhibited characteristics of “involutionary” innovation. Specifically, R&D investment has continued to rise, yet innovation outputs have become highly homogeneous, product differentiation has remained insufficient, and firms have struggled to form effective competitive advantages. This phenomenon not only undermines firms’ innovation efficiency but also distorts market competition mechanisms. Against this backdrop, whether the application of artificial intelligence (AI) can alleviate “involutionary” innovation and reshape innovation patterns has become an important research question. However, existing literature has largely focused on the impact of AI on firm-level performance, with limited attention paid to its influence on innovation behavior from an inter-firm interaction perspective. To address this gap, this paper examines how AI application affects “involutionary” innovation from the perspective of inter-firm innovation relationships and explores the underlying mechanisms. This study draws on data from Wind, CSMAR, the China Research Data Service Platform (CNRDS), the patent database of the China National Intellectual Property Administration, and the National Bureau of Statistics of China. Adopting the methodologies of Byun et al. (2021) and Shen Kunrong et al. (2023), it quantifies “involutionary” innovation by constructing patent similarity measures among firms. For the core explanatory variable, this study integrates AI-related keyword information extracted from corporate annual reports with relevant financial investment data and employs the entropy method to construct a composite index of firms’ AI application level, thereby providing a more comprehensive measure of digital technology adoption. The empirical results show that AI application significantly reduces patent similarity among firms, effectively alleviating “involutionary innovation.” Mechanism analysis reveals that AI alleviates “involutionary” innovation through three synergistic channels: enhancing human capital, promoting specialization, and increasing strategic differentiation. Further heterogeneity analysis indicates that the inhibitory effect is more pronounced in state-owned enterprises, labor-intensive firms, and firms operating in competitive and high-tech industries, highlighting the heterogeneous effects of AI empowerment under different factor endowments and market environments. In addition, extended analysis finds that “involutionary” innovation significantly intensifies price competition and induces price wars, suggesting that it not only fails to improve innovation quality but may also lead to inefficient competitive outcomes. Compared with existing studies, the innovations and contributions of this paper are mainly fourfold. First, this study moves beyond a single-firm perspective by incorporating inter-firm interactions and introducing patent similarity into the measurement of “involutionary innovation,” thereby extending the analytical dimensions of corporate innovation research. Second, this study integrates text mining with financial data to construct a comprehensive index of AI application, thereby improving measurement accuracy and robustness. Third, this study develops a “resources-capabilities-strategy” analytical framework to systematically uncover the mechanisms through which AI influences “involutionary innovation,” enriching the theoretical understanding of the relationship between AI and corporate innovation. Fourth, this study, from the perspective of competitive consequences, reveals the reinforcing effect of “involutionary” innovation on price competition, further expanding the scope of related research. This study demonstrates that AI serves not only as a key driver for enhancing firms’ innovation capabilities but also as a crucial force in alleviating “involutionary” innovation and optimizing competitive structures. However, its effects are subject to significant contextual constraints. Therefore, policymakers should avoid one-size-fits-all approaches and instead adopt targeted and differentiated policies. Meanwhile, firms should carefully promote AI adoption based on their resource endowments and development stages, in order to better unleash AI’s value in fostering differentiated innovation and enhancing competitive advantages.