Family physicians increasingly encounter large language models (LLMs) and other AI tools embedded in clinical, private, and patient-facing systems. These tools can support guideline-concordant, evidence-based communication and continuity of care for the experienced and trained physicians. They may also widen inequities and create confusion with misinformed usage or if outputs are not adapted for language, culture, literacy, disability, or digital access. This workshop presents a replicable, vendor and tool-neutral framework for using AI in clinical workflows that strengthen relational and equitable care in time-pressured practice.Participants in this workshop will be able to: Map a high-impact clinical scenario into an AI-supported consultation workflow. Design prompts and scripts that promote clinical ownership, transparency, and shared decision making. Apply a SAFE-AI framework (Scope, Accountability, Fairness, Explainability) to identify risk, bias, and communication gaps. Adapt AI-generated content for multilingual and multicultural populations. Use safe prompt-design ("prompt engineering") templates transferable across different AI tools/modalities.Short interactive plenary plus a live demonstration of an AI-supported consultation, followed by a guided small-group design sprint, using participants cases (e.g., multimorbidity, mental health, chronic disease management). Facilitators provide a structured SAFE-AI checklist, ready-to-use prompt templates, and documentation phrases to guide participants. Groups generate patient-friendly explanations, test them via brief role-play, then refine outputs with an equity lens and best-practice guardrails (verification, limits, and clinician sign-off). 0-10 min: Introduction and poll of AI usage. 10-20 min: Live demonstration. 20-40 min: Group workflow build using provided templates. 40-55 min: Peer simulation and feedback. 55-60 min: One-page AI use protocol and individual commitment to a practice pilot.Clinical AI-assisted workflow tailored to their needs SAFE-AI checklist and draft practice protocol Take-home prompt and workflow templates adaptable across clinical settingsGenerative AI is increasingly shaping how patients receive information and how clinicians manage cognitive and administrative load. This presentes a unique opportunity for family physicians to harness language models to deliver multicultural, multilingual, and equitable primary care while keeping, safe AI methods and clinical expertise central. By intentionally designing AI-augmented workflows and developing transferable skills in prompt design, participants can enhance relationship-based, evidence-informed care, improving patient understanding, engagement, and shared decision-making.