Good evening

A Message from the Innovation and Technology Department

AI tools such as ChatGPT, Claude, Gemini, and Microsoft Copilot are now widely available and many of us use them regularly. This is a reminder of how to use them safely and responsibly at APHI.

What AI is good for:

• Generating ideas, drafts, and suggestions
• Summarizing long documents
• Helping with writing, formatting, and translation
• Learning new concepts quickly

What AI cannot do:

• Understand your current, real-world situation
• Access our internal systems, live data, or company context (unless specifically set up to do so)
• Replace your professional judgment or decision-making
• Guarantee accuracy — AI can and does make mistakes, including confident-sounding wrong answers

Why you cannot use AI ideas at 100%

AI generates responses based on general patterns from its training data. It does not know your team, your patients, your current operations, your budget, or your actual constraints. This means:

• An AI-generated idea may sound logical but be completely unrealistic in our context
• Following AI output blindly can lead to wrong decisions, wasted resources, or real harm
• A response that is 100% from AI, without human review, is likely to be 100% incomplete — because AI simply does not have the full picture

Always treat AI output as a starting point, not a final answer. Take the idea, check it against reality, discuss it with your team, and then decide.

Important rules to follow:

1. Do not paste confidential company data into public AI tools. This includes patient information, financial records, contracts, or anything marked internal or confidential.
2. Do not rely on AI output as final. Always verify before acting, especially for clinical, financial, legal, or operational decisions.
3. AI does not know our current situation. It has no access to what is happening in our business today. Use it for ideas and starting points only.
4. Never present AI-generated content without reviewing it first. Check with your team or supervisor before acting on any AI suggestion.
5. You are responsible for any output you use. If you present AI-generated content as your own work, you own the accuracy of that content.
6. AI in the wrong hands is dangerous. Staff who are not trained or guided on AI use can cause serious problems — spreading wrong information, making poor decisions, or exposing confidential data — without even realizing it.
7. Do not believe AI 100%. AI can sound very confident while being completely wrong. Always apply your own knowledge, experience, and common sense.
8. When in doubt, ask. Contact the IT team if you are unsure whether a specific AI use is appropriate.

Used well, AI is a powerful assistant. Used carelessly, it can cause real harm. Think before you prompt. Verify before you act. And never let a machine make decisions that belong to a human.

For questions or guidance, contact the IT team.