Objective To evaluate the accuracy of automatic generation of surgical disease nursing diagnoses by different artificial intelligence (AI) tools, test their potential in generating nursing diagnoses, and expand the application scope of AI in the nursing field. Methods In December 2024, 10 typical cases from Surgical Nursing were selected.Unified instructions were input into 5 AI tools (Doubao, ChatGPT-4, Kimi, ERNIE Bot, and Qwen) respectively to generate surgical disease nursing diagnoses automatically.Eleven experts were invited to evaluate the quality of the generated nursing diagnoses. Results The quality scores of nursing diagnoses generated by the 5 AI tools, ranked from highest to lowest, were Doubao, ChatGPT-4, Kimi, ERNIE Bot, and Qwen; there was a statistically significant difference in the quality scores of surgical disease nursing diagnoses generated by different AI tools (P＜0.05).The Fleiss′s Kappa value among the 11 experts′ evaluations of the quality of 10 generated nursing diagnoses was 0.445 (P＜0.05). Conclusion The quality of surgical disease nursing diagnoses automatically generated by AI tools has high recognition, but further accuracy judgment combined with clinical experience is needed in practical use.