Side-channel analysis techniques extract cryptographic keys by analyzing physical or electrical characteristics generated during the encryption process. When performing correlation power analysis and simple power analysis on raw traces, challenges such as noise interference necessitate effective trace preprocessing–a task that traditionally relies on domain expertise, specialized tools, and extensive experience. Meanwhile, ChatGPT has gained widespread attention for its intelligent interaction capabilities and effectiveness in assisting users with task-specific operations. However, its potential for trace preprocessing remains underexplored and calls for systematic investigation. In this paper, we propose three expert-strategy prompt templates to explore and assess ChatGPT’s capabilities in trace preprocessing. We validate the effectiveness of ChatGPT in performing six categories of trace preprocessing methods through expert-strategy prompts at varying abstraction levels. Furthermore, we evaluate the impact of ChatGPT-assisted preprocessing on traces across various platforms and cryptographic algorithms, analyzing its influence on the overall performance of side-channel analysis.