1. Purpose and scope
This policy explains how AI is used on this website and in my client work, so you know what happens with the information you share. It covers the Ask box, the brief chat, the "Prefer your own AI?" links and the AI tools I use while working on projects.
2. AI on this site
- Ask box. It answers from this site's own pages. When AI answers are switched on, it uses Anthropic's Claude API, limited to the facts on this site. AI answers can be wrong: they are a starting point, not advice or a quote.
- Brief chat. A short scripted set of questions, not an AI. It fills in the project form in your browser and never sends anything on its own.
- Prefer your own AI? Opens ChatGPT or Claude in your own account with a ready prompt. That service's terms apply to what you type there.
3. AI in client work
I use AI tools as assistants: for drafts, research, code suggestions and checking work. I review and test everything before it reaches a client, and I stay accountable for the result, not the tool.
4. Handling client information
Allowed
- Public material, such as published pages and public documentation.
- Project material with names and personal details taken out.
Not allowed
- Passwords, API keys, access tokens and other access details.
- Personal data of a client's customers or staff.
- Anything a client has marked as confidential, unless the client agrees in writing.
5. Results and feedback
AI does not write client feedback, results or case study numbers. Feedback is quoted as clients wrote it, and results come from the real projects.
6. Your choice
If you want no AI tools used on your project, or only specific ones, tell me before we start and I will work that way.
7. Changes and contact
I may update this policy as tools and practice change. The date at the top shows the current version. Questions? Send a note through the form on the contact page or message me on LinkedIn.