Responsible AI Use Policy

1. Work accounts only

We use AI tools only through business accounts whose terms we have read and saved. We never use personal chatbot accounts for any client work.

2. What never goes into an AI tool

• Personal information about a client's members or employees (names, status, community, hours, pay, certifications), unless the client has approved that specific tool in writing.
• Client rate sheets, contracts, or financial records, outside the client's own approved tools.
• Passwords, keys or access credentials of any kind.

3. AI drafts, people decide

AI may draft documents, summaries and checks. A named person approves anything that is sent to a client, a regulator or an operator, and anything that affects a person: invoices, reports, safety forms, hiring and scheduling decisions.

4. No decisions about people by AI

We do not build systems that score, rank, discipline or monitor individual workers. Safety tools we build support crews and the qualified person who signs off; they do not replace that person.

5. Every number traceable

Any figure an AI tool produces in our work must be traceable to a source document. A figure that cannot be traced is removed, not guessed.

6. Demonstrations use synthetic data

Every demo, test and training example we show publicly uses invented data and is labelled as synthetic.

7. Training and settings

Where a tool allows it, we switch off use of our data for model training. We review every AI tool's terms each quarter and update our public tool list when anything changes.