BYU School of Medicine
On this page
A Guide for Medical Students

Mastering AI in Medicine

"Innovations such as artificial intelligence [can] both (1) assist you in receiving magnificent blessings and (2) diminish and suffocate your moral agency. Please do not allow the supposed accuracy, speed, and ease of modern technologies to entice you to avoid or circumvent the righteous work that invites into your life the blessings you will need."
— Elder David A. Bednar, BYU Devotional, January 2024

Understanding how to wield AI effectively isn't just a tech skill — it's becoming a clinical competency. This guide will transform how you learn, study, and prepare for your calling as a healer.

Begin
Part 1

Getting Started with AI

Why does AI give me generic answers?

Here's the truth most people never realize: AI is a mirror. Vague questions get vague answers. Specific, context-rich questions get expert-level responses.

❌ What most students type
"What causes chest pain?"
✓ Effective prompting
"I'm a second-year medical student studying for Step 1. Explain the differential diagnosis for acute chest pain in a 55-year-old male with diabetes and hypertension, organized by most likely etiology. Include the key distinguishing features for each."

The second prompt gives AI the context to tailor its response to your level, your purpose, and your specific scenario.

💡 The Rule: Treat AI like a brilliant colleague who just walked into the room. They're incredibly capable, but they don't know your situation yet. The more context you provide, the more useful they become.

Part 2

AI Models & Access

Which tools to use and when

Not all AI is created equal. Here's what you need to know:

General-Purpose AI Models

Use these for general studying, concept explanations, and basic questions:

Model Access Best For
ChatGPT chat.openai.com General questions, explanations, study help
Claude claude.ai Nuanced analysis, longer documents, careful reasoning
Gemini gemini.google.com Google integration, multimodal tasks
Grok x.com/i/grok Real-time information, current events

Medical-Specific AI Tools

For clinical questions, case analysis, and medical decision support — use these specialized tools:

Tool Access Best For
OpenEvidence openevidence.com Evidence-based clinical questions, citations included
Amboss amboss.com Medical knowledge, Step prep, clinical reasoning
UpToDate uptodate.com Clinical decision support, treatment guidelines

💡 When to use which: General inquiries and basic study questions → use ChatGPT, Claude, Gemini, or Grok. Medical cases and clinical questions → use OpenEvidence, Amboss, or UpToDate for more detailed and precise answers with proper citations.

Your AI Access for This Course

🌐 Cloud Models — Via This Website: You have access to state-of-the-art AI models (GPT-4o, Claude, and more) directly through the AI Chat on this website. Use these models at no cost for your studies.

🎓 Set Up Your Own Access: Consider creating free accounts with tools like OpenEvidence for personalized medical tutoring. Having your own accounts lets you build a history of interactions, get tailored recommendations, and continue learning beyond this course.

Part 3

Privacy & HIPAA

The one rule you must follow

🚨 The Rule: NEVER put patient data (PHI) directly into any AI model. Period.

This applies to all AI tools — ChatGPT, Claude, Gemini, Grok, and any other consumer AI. These tools are NOT HIPAA compliant.

What counts as PHI?

What CAN you do?

💡 Bottom line: When in doubt, don't include it. Use AI to learn concepts and explore ideas — never to process real patient information. This may change as HIPAA-compliant AI tools become available in clinical settings, but for now: no patient data in AI.

Part 4

Getting Better Results

Techniques for more useful AI responses

The Context Window

The "context window" is how much information AI can hold in memory during a conversation. Think of it as working memory — the bigger it is, the more information AI can reference at once.

How to use this strategically:

  1. Paste in source material — Give the AI your lecture notes, a paper abstract, or detailed case information
  2. Keep important info near the end — AI pays more attention to recent messages
  3. Start fresh for new topics — Old context can confuse new questions
⚡ Try This
"Here are my notes from today's cardiology lecture: [paste notes]. Now quiz me on the key concepts with clinical vignette-style questions similar to Step 1."

You just created a personalized study tool that knows exactly what you learned today.

Feed It Sources, Get Better Answers

AI models have knowledge cutoffs and can "hallucinate" — confidently stating incorrect information. The solution? Give it the source material.

❌ Instead of asking
"What's the latest treatment for HFrEF?"
✓ Try this
"Based on the 2023 ACC/AHA heart failure guidelines (which I'll paste below), summarize the first-line treatments for HFrEF and explain the evidence level for each. [paste guideline excerpt]"

Now the AI is synthesizing real, current information — not guessing from training data that might be outdated.

Sources to feed AI:

  • Lecture slides or transcripts
  • UpToDate summaries
  • Paper abstracts or full text
  • Clinical guidelines
  • Your own notes
Part 5

Critical Thinking

Using AI responsibly and effectively

Always Verify

AI is confident. That doesn't mean it's correct.

AI will never say "I don't know" with uncertainty in its voice. It delivers wrong answers with the same confidence as right ones. This is one of its most dangerous characteristics.

Your verification toolkit:

The more consequential the information, the more important verification becomes. Studying for an exam? Verify key facts. Real patient care? Never rely on AI alone.

"Artificial intelligence cannot replace revelation or generate truth from God. We have the responsibility to ensure the Holy Ghost can attest to the truth and authenticity of all we say and share."
Elder Gerrit W. Gong, BYU Education Week, August 2025 (source)

Common Misconceptions

Myth Reality
"AI will give me the right answer" AI gives you an answer. Correctness is your job to verify.
"Longer prompts are annoying" AI loves context. More detail = dramatically better responses.
"AI is cheating" AI is a tool, like a calculator. Using it wisely is a skill worth developing.
"I can just copy AI responses" AI is a drafting partner. Your judgment must refine the output.
"All AI chatbots are basically the same" Models vary wildly in capability. Choosing the right tool matters.
"AI knows the latest research" AI has knowledge cutoffs. For recent info, provide sources yourself.

When NOT to Use AI

Knowing when to step away is as important as knowing how to use it.

AI is powerful, but there are times when reaching for it is the wrong choice:

🚫
During proctored exams

This should be obvious, but using AI during exams is academic dishonesty. Beyond the ethical violation, it also means you won't actually learn — and that catches up with you on the wards.

🚫
For anything involving real patient PHI

Never paste patient information into AI tools. Most are not HIPAA compliant. De-identify completely or create fictional scenarios based on the clinical pattern.

🚫
When you haven't tried first

If you immediately reach for AI without attempting to reason through a problem yourself, you're robbing yourself of the productive struggle that builds clinical thinking. Try first, then use AI to check, expand, or clarify.

🚫
For documentation requiring your authentic voice

Personal statements, reflective essays, and certain assignments are meant to capture YOUR thinking and growth. Using AI undermines their purpose and your development.

🚫
When the stakes require human judgment

Real patient care decisions, sensitive conversations, ethical dilemmas — these require human wisdom, empathy, and accountability that AI cannot provide.

The goal is to become a capable physician who can use AI as a tool — not a dependent user who can't function without it.

"AI cannot replace the gift of divine inspiration or the individual work required to receive it. Interactions with AI cannot substitute for meaningful relationships with God and others."
General Handbook, December 2025 (source)
Part 6

Quick Reference

Checklists and guidelines for responsible AI use

Best Practices Checklist

✦ Before You Prompt
  • What's my specific goal with this interaction?
  • What context does AI need to help me well?
  • Am I using the right model for this task?
✦ In Your Prompt
  • State your role and level (medical student, studying for X)
  • Describe the specific scenario or question
  • Explain what format you want (bullets, table, explanation)
  • Paste relevant source material when available
✦ After You Get a Response
  • Does this actually answer my question?
  • Have I verified the critical facts?
  • Can I explain this in my own words now?

Responsible Use & Citation

AAMC principles summary: Use AI to augment learning (not replace your judgment), protect privacy and confidentiality, evaluate outputs for bias and inaccuracy, and disclose when AI contributed to your work. Read the AAMC Principles for AI Use in Medical Education →

APA citation format for AI

When required by your assignment, cite AI assistance clearly. Example:

OpenAI. (2026). ChatGPT [Large language model]. https://chat.openai.com
Disclosure best practices
  • Briefly disclose how AI was used (brainstorming, outlining, explanation support, etc.).
  • Identify what you verified independently with course materials or primary sources.
  • Keep authorship and accountability with you — revise in your own voice and reasoning.
  • Never enter identifiable patient information into consumer AI tools.

The Future Is Collaborative

The best physicians of tomorrow will be those who know how to learn, synthesize, and leverage every tool available to serve their patients. AI is an incredibly powerful tool. Use it wisely. The goal is to amplify your thinking, not outsource it.

Elder Gong has taught that these technologies are "part of the Lord hastening His work in the latter days." You have the opportunity to use AI to serve more patients, catch diagnoses faster, and extend care to those who need it most.

BYU School of Medicine Mission Statement

The Brigham Young University School of Medicine is founded on the teachings and example of Jesus Christ, the Master Healer, and guided by the doctrine and leaders of The Church of Jesus Christ of Latter-day Saints.

We provide a spiritually grounded and scientifically rigorous education for physicians who will minister to God's children throughout the world as healers, teachers, researchers, leaders, and disciples of Jesus Christ.

We approach the practice of medicine as an opportunity to prevent and alleviate human suffering and to care for all people as fellow children of a loving God.

BYU School of Medicine Mission Statement →