Carpal tunnel syndrome (CTS), a common peripheral neuropathy impacting approximately 3–6% of the general population with greater prevalence in individuals over 50 years of age, causes pain, numbness, and tingling in the hand and wrist, typically due to compression of the median nerve at the wrist [[1], [2], [3]]. The severity of CTS can range from mild symptoms that resolve with conservative measures to more severe cases requiring surgical intervention to prevent permanent nerve damage [1]. Treatment options for CTS includes wearing a neutral wrist splint at night or taking oral medications like nonsteroidal anti-inflammatory drugs, vitamin B12, and pregabalin, which have proven effective for CTS [4]. Surgery for CTS may be necessary in cases where there is thenar muscle atrophy or when conservative treatments do not provide relief [4].

In 2016, the American Academy of Orthopaedic Surgeons (AAOS) released the Appropriate Use Criteria (AUC) for the management of carpal tunnel syndrome to guide clinicians in making evidence-based treatment decisions [5]. However, the increasing use of artificial intelligence (AI) in healthcare has prompted new questions regarding the potential role of AI tools in clinical decision-making for CTS management. The two most popular AI tools, Chat Generative Pre-trained Transformer-4 (ChatGPT-4o, OpenAI, San Francisco, CA), and Google’s Gemini 1.5 Pro chatbot (Google LLC, Mountain View, CA, USA), have shown promise in assisting with a wide range of medical inquiries, from preliminary diagnoses to treatment recommendations, though its accuracy in specific applications, such as CTS, has not been fully evaluated [[6], [7], [8]].

Recent studies suggest that while ChatGPT-4o can provide high-quality general information, its ability to deliver personalized treatment recommendations for CTS—based on varying degrees of symptom severity and individual patient characteristics—remains uncertain. This analysis is further enhanced by comparing ChatGPT-4o’s recommendations to those generated by Gemini 1.5 Pro, especially given Gemini 1.5 Pro’s appearance in Google searches as a primary source of information [8]. While ChatGPT has been shown to deliver valuable educational content, it falls short in offering clinically accurate or nuanced management plans compared to human experts [9,10]. As AI tools like ChatGPT and Gemini 1.5 Pro gain more popularity in medical practice, evaluating their potential utility in guiding the management of conditions like CTS is critical.

The primary aim of this study was to assess how accurately ChatGPT-4o and Gemini 1.5 Pro recommend treatment modalities for carpal tunnel syndrome based on patient scenarios. In evaluating the AI recommendations against established guidelines, such as the AAOS AUC for Carpal Tunnel Syndrome, the study sought to establish which of these two leading AI platforms would offer greater consistency with recommended treatments, ensuring that accessible AI tools align with established medical guidelines.