Generative artificial intelligence has witnessed explosive growth in recent years. Large language models powered by AI are increasingly adopted by researchers to assist in scientific research and academic writing. While these tools help reduce the time investment required from researchers and significantly enhance research efficiency, they simultaneously raise various academic ethical concerns, including intellectual property infringement and academic misconduct. In response to AI's impact on academic research and publishing, institutions worldwide have established guidelines governing AI tool usage for researchers and academic publications. This paper analyzes policy statements regarding AI tool usage issued by 50 Chinese social science academic journals since 2023, representing diverse disciplines and institutional affiliations, to comprehensively understand the awareness and responses within China’s social science academic journal community regarding AI tools, and provide evidence-based insights for improving AI policies in academic publishing. Through detailed textual analysis, the research identified nine key aspects addressed in AI policies of academic journals : responsibility definition, prohibited actions, permissible scope, disclosure requirements, verification mechanisms, penalty measures, multiple stakeholders, dynamic adjustments, and an open attitude. These interconnected and organically unified elements form a cohesive AI Usage Policy framework for academic social sciences journals. They delineate the boundaries for authors to reasonably use AI tools for academic innovation and provide fertile ground for academic prosperity in the AI era, while charting a future vision for the development of academic journals themselves. findings indicate that the domestic academic journal community demonstrate sufficient awareness of AI tools and have responded promptly, maintaining a balance between regulatory oversight and supportive implementation while showing a trend toward dynamic policy optimization. However, several limitations exist in the existing AI policies of academic journals: most journals lack specific AI usage guidelines, some journal policies lack operational practicality, some journals with established AI policies suffer from inadequate implementation, and existing AIGC detection tools require improvement. Addressing AI-related challenges in academic publishing first requires an open and inclusive attitude, acknowledging technological advancement's inevitability. Second, it is essential to adhere to the principles of prudent responsibility, guiding technological development through the establishment of rules that define AI usage boundaries while preserving space for development, ultimately fostering a positive interaction and synergistic development between AI technology and academic publishing. Five specific measures are recommended: promoting the standardization of AI usage guidelines in academic publishing, leveraging AI to advance academic innovation and disciplinary development, fully unleashing the potential of AI tools to empower journal editors, regulating AI involvement in academic creation at its source, and establishing a robust training and guidance system for AI applications in research and editing.