LTTR's tagline is "Type Letters. Own Your Data." Our privacy promise is clear: no AI reads your letters, no LTTR cloud stores them, no one trains a model on your words.
But here's something we want to be transparent about: we build LTTR using AI-powered development tools.
We're saying this upfront, not in fine print. Here's why — and what it means for your letters.
What "No AI" means in LTTR
When LTTR says your letters are private from AI, it means exactly that:
No AI reads your writing. Not OpenAI, not Anthropic, not Google, not anyone.
No training on your content. Your letters never reach an AI service as training data.
No hidden processing. There are no "smart suggestions," "AI improvements," or background analysis of your words.
No LTTR servers storing your letters. The local-first architecture means letters stay on your device until you export them. Cloud AI can't reach what never leaves.
This is the promise. It doesn't change.
How we build the app
LTTR is built by a solo developer. Like the vast majority of modern software teams — from startups to large engineering organizations — we use AI-powered coding tools in our development workflow:
AI code editors and completion tools to write boilerplate faster
AI debugging assistants to diagnose problems during development
AI documentation tools to generate code documentation
These tools see LTTR's source code, not your letters. The two are completely separate.
Why use these tools?
Honestly, building a full-featured, privacy-respecting iOS app as a solo developer is hard. AI development tools let one person ship what would otherwise require a team. That means more time focused on the things that matter most: the privacy architecture, the export quality, and the writing experience.
The faster we can build, the sooner users get the features they've asked for. The better our debugging tools, the fewer bugs reach your device.
The distinction that matters
There's a meaningful difference between:
Using AI to build software → Developer productivity tool, never sees user data
Using AI to surveil users → AI reads, analyzes, or trains on user content
LTTR does the first. Not the second.
This distinction will matter more over the next few years, as AI becomes embedded in nearly every development workflow. We'd rather be honest about it than pretend AI tools don't exist in our stack, and then have that discovered later and feel like a betrayal.
Privacy-conscious users deserve the full picture. That's what transparency means.
What about future AI features?
We get asked occasionally about adding AI features to LTTR.
Our position: if optional AI features are ever added, they will:
Be opt-in only. Default off. You choose to enable them.
Process on your device. No sending your letter text to a cloud AI service.
Be fully disclosed. We'll explain exactly what data leaves your device (if any), before you enable anything.
The local-first architecture makes cloud AI difficult to integrate anyway, which is partly by design. Your letters staying on your device isn't just a marketing claim. It's a structural constraint that protects you even from things we haven't thought of yet.
Our commitment
LTTR's value is built on trust. Radical transparency about what the app does, what it doesn't do, and how it's built is part of keeping that trust intact.
If this changes, we'll tell you. If we add AI features, we'll tell you before you enable them. If our tools or architecture evolve, we'll update this post.
Your letters belong to you. That promise doesn't have fine print.
Try LTTR: mautonce.com/lttr
Privacy policy & terms: mautonce.com/lttr#policy-and-terms
Concerns, Questions, or Suggestions: LTTR@mautonce.com
Published at launch. Last updated: May 2026.