The Handshake: My AI Framework

I work with AI, I do not let AI work for me. Over time, I have had the opportunity to engage with it in many meaningful ways. Looking ahead, I see a future where healthcare and AI work hand in hand. However, before that partnership can take shape, both sides have important things to work through — before the agreement, before the handshake.

How am I going to use what I have learned in healthcare? I start by acknowledging the barriers that come with working alongside AI. The most significant barrier I encountered is the tension between AI use and the Health Insurance Portability and Accountability Act (HIPAA). This required me to find creative, responsible ways to protect patient privacy while still achieving my goal.

My goal is to ensure that all signs and symptoms are thoroughly assessed, and that diagnoses are accurately assigned with the support of patient data. AI can help streamline administrative responsibilities — such as completing medical records — while freeing the clinician to focus on the patient.

Critically, this does not replace the healthcare provider’s role in assessing, evaluating, diagnosing, and caring for the patient. In this model, AI functions as an assistant , not a fellow provider. This distinction preserves time and space for genuine human connection with patients. I also believe that AI usage (and what it is used for) should always be disclosed, so patients understand what operates in the background. Transparency builds trust.

The Human–AI Partnership Model

The Human–AI Partnership Model centers on one idea: using patient information to accurately complete medical reports, while confirming that signs and symptoms are captured with the highest-probability diagnoses surfaced to support the provider’s clinical reasoning , never to replace it.

Step 1: Defining the Problem

Due to reasons like HIPAA violations and possible incorrect information, AI has usage in healthcare has been a bit discouraged.

Step 2: Exploring With AI

Using Claude, I brainstormed concrete strategies for curbing HIPAA violations while still harnessing AI’s capabilities. The conversation helped me map potential solutions I had not considered on my own. Below is a screenshot of what we developed together:

Step 3: Validating Evidence

I reviewed peer-reviewed research, professional guidelines, and trusted resources that explored AI usage in healthcare in depth. This process surfaced important considerations around algorithmic bias, clinical integration pointers, and patient perceptions of AI-assisted care.

Step 4: Applying Clinical Judgment

Drawing on my nursing expertise and clinical experience, I reviewed all identified barriers and potential accomplishments. I reflected on how a provider like myself and an AI system could genuinely work together, each contributing what the other cannot. Clinical judgment cannot be outsourced but it can be supported.

Step 5: Reflect and Improve

Reflection is not a final step , it is an ongoing practice. Each interaction with AI is an opportunity to ask: Did this serve the patient? Did this uphold my clinical standards? Did this protect privacy? Continuous improvement is built into the model, not added on afterward.

Ethical Safeguards Within the Framework

• Clinician Review: Before deploying any AI-assisted assessment, I verify its output against my clinical judgment and current best practice guidelines to ensure accuracy and patient safety.
• Symptom Capture: I confirm that all presenting symptoms are documented in full — regardless of their perceived relevance to the final diagnosis — so that nothing is overlooked.
• Relationship Mapping: Each identified symptom is connected back to the primary clinical concern, and its importance is triaged to support clear, prioritized reasoning.
• Access Control: Before any information is entered into an AI system, I ensure that all patient-identifying details are removed, limiting data to de-identified inputs used only within secure, internal workflows.

The Handshake

A handshake is an agreement between two parties — a moment of mutual recognition, trust, and shared purpose. That is precisely what this framework represents. Human clinical judgment and AI support do not compete; they complement. The provider brings irreplaceable empathy, ethical accountability, and contextual wisdom. AI brings speed, pattern recognition, and the ability to surface what might otherwise be missed. When both honor their role, the patient is better served. This is the handshake: not a transfer of authority, but a partnership built on transparency, integrity, and an unwavering commitment to care.