Background: Artificial Intelligence (AI) has become increasingly integrated into software development, with tools such as ChatGPT, GitHub Copilot, Cursor AI, Claude, DeepSeek, and Amazon Code Whisperer now widely used to assist developers. While previous research has highlighted the potential of these tools to improve individual productivity, their influence on Agile software development practices has received comparatively less attention.

Aims: This study examines how AI-assisted coding tools affect Agile software development practices, focusing on developer productivity, code quality, Agile workflows, and associated challenges.

Methods: A quantitative survey was conducted with 91 software professionals using AI-assisted coding tools in Agile environments. Responses were analyzed using descriptive statistics, reliability testing (Cronbach's Alpha), and correlation analysis.

Results: AI-assisted coding tools are widely perceived as beneficial. Participants reported improvements in coding efficiency (M = 4.20, SD = 0.68), faster task completion (M = 4.37, SD = 0.77), and better support for Agile activities such as requirement clarification (M = 4.08, SD = 0.70) and sprint execution (M = 4.00, SD = 0.67). However, concerns were raised regarding over-dependence (M = 3.79, SD = 0.75), security issues (M = 3.78, SD = 0.74), and reduced learning opportunities (M = 4.26, SD = 0.99).

Conclusion: The study indicates that AI-assisted coding tools can provide meaningful support for Agile development when used alongside human expertise. Organizations should adopt them carefully and establish practices that encourage responsible use, continuous learning, and quality assurance