How we use AI

Last updated: August 13, 2026

I want to be upfront about how AI is used at LazyFinance. There are two separate things: the AI that helps me build the app, and the optional AI assistant you can turn on inside the app. Here is exactly what each one sees, and what neither one sees.

How AI helps build LazyFinance

AI coding assistants help me write and refactor code more quickly. But it's a tool, not the decision-maker: I review what it produces, I test it, and I'm responsible for everything that ships. The product ideas, the architecture, and the security model are mine.

And the cute sloth mascot? That's not AI either: my sister drew it by hand.

The development tools never touch your data

Development AI is used on my machine and in my tools. It is never connected to the running app and never has access to your financial data or your account. And the Insights page stays deterministic: plain, auditable rules computed from your own numbers, no AI model involved.

The optional assistant

The assistant, part of the LazyFinance AI plan, lets you ask questions about your own finances. It is strictly opt-in: nothing is sent to any AI service until you explicitly turn it on, and if you never turn it on, nothing changes for you.

If you do turn it on, here is exactly how it works:

• When you ask a question, your message and the data needed to answer it (account balances, transactions, categories, merchants, bills, insights) are processed by Anthropic, the company behind the Claude model, on servers in the United States.
• Anthropic does not use your data to train its models, and keeps it for at most about thirty days for abuse monitoring before deleting it.
• You control what the assistant can see: transaction notes stay private unless you share them, and you can hide any category so its transactions never appear anywhere in its answers, totals included.
• The assistant can also prepare changes: adding, fixing, or deleting transactions, or updating a balance. Nothing is applied until you confirm the card shown in the chat: every change needs your explicit OK, and you see exactly what will happen before you click.
• Your conversations are kept with us for at most 90 days. Clearing the conversation or turning the assistant off deletes them immediately.

The privacy side is spelled out in the Privacy Policy.

Who's in the driver's seat

I've been building and shipping software for years; AI speeds up the typing, it doesn't replace judgment. Architecture, data handling, and security decisions are made deliberately by me, reviewed line by line, and tested, rather than left to whatever an AI happened to suggest. That's the difference between using AI as leverage and shipping unreviewed code.

Questions

Reach me through the Contact page. I read every message.