How I use AI, and how I handle your data
If I'm going to help your firm use AI responsibly, you should be able to see how I use it in mine.
The rules I work to
No personal information from you or your clients goes into a public or free AI tool.
I only use AI tools I've vetted for data location and training use, under a data processing agreement.
Anything AI produces for a client is read and checked by me before you see it.
I don't use AI to give tax, legal or financial advice, and nothing I build for you does either.
The tools I use
Claude, for writing, analysis and building software, with the setting that lets Anthropic train on my conversations switched off.
Google Workspace, for email, documents, file storage, forms and bookings.
Make.com, for automations.
Airtable, for tracking work.
Netlify, for hosting websites.
Stripe, for card payments.
When I work with your data
Your data
If a piece of work means handling your clients' personal data, you're the controller and I'm your processor. We sign a data processing agreement before anything starts. Tools and accounts are set up in your name wherever possible, so the data stays in systems you control.
When I build AI into your firm
Everything follows a written policy you approve. Tools go on an approved list only after the same checks I apply to my own. A person at your firm reviews anything client-facing before it goes out, and use is logged so you can show what was done and who checked it.
If something goes wrong
If I become aware of a problem affecting your data, I'll tell you promptly and help you deal with it, including anything you need to report.