Not on the accounts we use. The business terms say so in plain words. Anthropic: Anthropic may not train models on Customer Content from Services. Microsoft Azure: prompts and outputs are NOT used to train any generative AI foundation models without your permission or instruction. We never build on free personal accounts, where that promise doesn't hold. Source: each provider's own terms.

We're not your lawyers, and we'll say so. What we do is build so your duties are easy to meet: your data stored in the region you choose, a written data agreement with us, and a person signing off where the law expects one (California's rules on automated decisions, Quebec's Law 25).

Answers show where they came from, so anyone can check. And nothing it writes reaches a customer until a person has read it.

Only what you choose to let it do. Reading and drafting, yes. Sending, paying or deleting waits for a person on your side.in a live build: a paid-ads system reads the ad accounts read-only. It can recommend a change. It can't make one.

It's how you stop it. We give the team one approved tool on your own accounts, each person with their own access, so nobody needs a personal chatbot for company work.

We work remotely, from Cape Town, as independent contractors, never as employees. You get an invoice in dollars, euros or pounds, paid by bank transfer or card. Our working day overlaps with Europe's and with US East Coast mornings.

Yes. The Vision shows you what it would look like, how we'd build it and what it costs. If you don't go ahead, you keep it.