Conversational Artificial Intelligence in Primary Care: Clinical Applications, Safety Considerations, and Future Directions - A Narrative Review

Background: Rapid advances in artificial intelligence (AI), particularly conversational systems powered by natural language processing (NLP) and large language models (LLMs), are transforming healthcare delivery. In primary care, these tools may help address workforce shortages and increasing patient demand. Objective: This review aims to critically evaluate the classification, clinical applications, effectiveness, safety concerns, and future potential of AI-driven chatbots in general medical practice. Methods: A narrative review of literature published between January 2020 and March 2026 was conducted using PubMed, Scopus, Web of Science, and Google Scholar. Eligible sources included randomized controlled trials, observational studies, systematic reviews, and policy papers. Evidence was synthesized thematically. Results: Chatbots were categorized into rule-based, AI-driven conversational agents, hybrid systems, and domain-specific tools. Their applications include symptom triage, patient education, chronic disease management, mental health support, and clinical documentation. While these technologies enhance accessibility and reduce administrative workload, concerns persist regarding diagnostic accuracy, algorithmic bias, hallucination errors, data privacy, and medico-legal accountability. Conclusion: Conversational AI has significant potential to augment primary care services; however, it cannot replace clinical judgment. Safe integration requires rigorous validation, ethical safeguards, and robust regulatory frameworks. Future research should prioritize real-world effectiveness and equitable implementation.