Abstract Introduction Artificial intelligence (AI) tools such as ChatGPT and Grok are increasingly accessed by men seeking guidance on sexual health and support for erectile dysfunction (ED), a condition often accompanied by anxiety, low self-esteem, and relationship stress. As these tools become more common sources of health information, understanding their ability to address the psychological and emotional aspects of ED is crucial for ensuring accurate, safe, and empathetic communication. Objective This study aimed to assess the quality and reliability of AI-generated responses on the psychological aspects of erectile dysfunction and to compare ChatGPT and Grok in terms of information accuracy, safety, empathy, and the effectiveness of their guidance in addressing users’ concerns. Methods Fifty standardized questions addressing the mental health dimensions of male sexual dysfunction were presented to ChatGPT (version 5 omni, OpenAI) and Grok (xAI, web version). The questions covered ten domains, including stress, performance anxiety, depression, relationships, and coping strategies. Responses were independently evaluated by three urologists from three different institutions. Information quality was assessed using the DISCERN instrument (1–5), and understandability and actionability were measured using the Patient Education Materials Assessment Tool for Printable Materials (PEMAT-P; 0–100%). Misinformation was rated from 1 (no misinformation) to 5 (high misinformation), and safety from 0 to 2. Additionally, ten clinical case vignettes were analyzed to assess emotional awareness, core issue identification, empathy, and quality of guidance (each 0–2; total score = 10). Statistical and graphical analyses were performed using RStudio (version 2023.12.1+402). Results Across 50 questions, Grok scored higher on information quality (DISCERN 4.09 vs 3.65, p 0.001) and understandability (85.9% vs 83.0%, p = 0.004). Actionability was similar between models (65.9% vs 67.5%, p = 0.63). ChatGPT demonstrated lower misinformation (1.74 vs 1.87, p = 0.04) and higher safety scores (1.61 vs 1.07, p 0.001).In qualitative analyses of ten clinical case vignettes, no significant differences were observed between ChatGPT and Grok in identification of the core issue (mean 1.82 vs 1.79, p = 0.68), recognition of emotional distress (1.38 vs 1.35, p = 0.58), safety and risk awareness (1.32 vs 1.39, p = 0.43), quality of guidance (1.41 vs 1.32, p = 0.87), empathy and tone (1.52 vs 1.28, p = 0.12), or overall performance (7.49 vs 7.20, p = 0.51). Conclusions AI chatbots show promise as adjunct tools for addressing the psychological dimensions of erectile dysfunction. Grok delivered superior information quality and readability, while ChatGPT demonstrated greater safety and accuracy. Together, their complementary strengths highlight the emerging role of AI-assisted communication in enhancing patient education, emotional support, and holistic care for men with sexual dysfunction. Disclosure No
Soares et al. (Mon,) studied this question.