This study explores the potential of generative AI, specifically GPT-4, in enhancing pharmacogenomic (PGx) testing interpretation and supporting clinical genetics. The AI assistant was developed using Retrieval Augmented Generation (RAG) with a knowledge base containing Clinical Pharmacogenetics Implementation Consortium (CPIC) data. The system combines retrieval and generative techniques to provide context-aware, accurate responses to user queries. The AI assistant demonstrated high efficacy in addressing PGx-related questions, outperforming ChatGPT 3.5, especially in provider-specific queries requiring specialized data and citations. Key areas for improvement include enhancing accuracy, relevancy, and language clarity in responses. The integration of context-aware GPT-4 with RAG significantly improved the AI assistant's utility. RAG's ability to incorporate domain-specific CPIC data, including recent literature, proved beneficial. Challenges remain, such as the need for specialized genetic/PGx models to improve accuracy and the need to address ethical, regulatory, and safety concerns. The AI assistant was evaluated against a specialized PGx question catalog and compared with ChatGPT 3.5. Expert panels assessed the AI assistant's responses on accuracy, relevancy, risk management, language clarity, bias neutrality, empathetic sensitivity, citation support, and hallucination limitation. The AI assistant outperformed ChatGPT 3.5 in provider-focused queries, achieving 85% effectiveness compared to 69%. For patient/layperson queries, the AI assistant showed marginal improvements, achieving 82% effectiveness compared to 78%. The AI assistant's performance was enhanced through prompt engineering, which improved the accuracy, safety, and comprehensibility of responses. The system was designed to provide clear, patient-friendly information at a 6th to 7th grade reading level, ensuring accessibility and understanding. The AI assistant also incorporated guardrails to ensure responses were not interpreted as medical advice and to avoid potential harm. The study highlights the potential of generative AI in transforming healthcare provider support and patient access to complex pharmacogenomic information. While careful implementation is necessary, generative AI can significantly improve understanding of pharmacogenomic data. The AI assistant demonstrated reliability in clinical communication, with strengths in risk assessment, language, and low incidence of hallucinations. However, challenges remain in addressing a broad spectrum of patient inquiries and improving accuracy and relevance. The study underscores the need for further development of specialized biomedical language models and the importance of ethical considerations and regulatory frameworks in AI deployment in healthcare.This study explores the potential of generative AI, specifically GPT-4, in enhancing pharmacogenomic (PGx) testing interpretation and supporting clinical genetics. The AI assistant was developed using Retrieval Augmented Generation (RAG) with a knowledge base containing Clinical Pharmacogenetics Implementation Consortium (CPIC) data. The system combines retrieval and generative techniques to provide context-aware, accurate responses to user queries. The AI assistant demonstrated high efficacy in addressing PGx-related questions, outperforming ChatGPT 3.5, especially in provider-specific queries requiring specialized data and citations. Key areas for improvement include enhancing accuracy, relevancy, and language clarity in responses. The integration of context-aware GPT-4 with RAG significantly improved the AI assistant's utility. RAG's ability to incorporate domain-specific CPIC data, including recent literature, proved beneficial. Challenges remain, such as the need for specialized genetic/PGx models to improve accuracy and the need to address ethical, regulatory, and safety concerns. The AI assistant was evaluated against a specialized PGx question catalog and compared with ChatGPT 3.5. Expert panels assessed the AI assistant's responses on accuracy, relevancy, risk management, language clarity, bias neutrality, empathetic sensitivity, citation support, and hallucination limitation. The AI assistant outperformed ChatGPT 3.5 in provider-focused queries, achieving 85% effectiveness compared to 69%. For patient/layperson queries, the AI assistant showed marginal improvements, achieving 82% effectiveness compared to 78%. The AI assistant's performance was enhanced through prompt engineering, which improved the accuracy, safety, and comprehensibility of responses. The system was designed to provide clear, patient-friendly information at a 6th to 7th grade reading level, ensuring accessibility and understanding. The AI assistant also incorporated guardrails to ensure responses were not interpreted as medical advice and to avoid potential harm. The study highlights the potential of generative AI in transforming healthcare provider support and patient access to complex pharmacogenomic information. While careful implementation is necessary, generative AI can significantly improve understanding of pharmacogenomic data. The AI assistant demonstrated reliability in clinical communication, with strengths in risk assessment, language, and low incidence of hallucinations. However, challenges remain in addressing a broad spectrum of patient inquiries and improving accuracy and relevance. The study underscores the need for further development of specialized biomedical language models and the importance of ethical considerations and regulatory frameworks in AI deployment in healthcare.