AI Disclosure

AI Disclosure

How HonestFit uses AI to tailor applications, and the guarantees we give you about what it can and cannot say.

1. Verified facts only

HonestFit can only use facts you have verified in your Facts Ledger. It never invents skills, titles, years of experience, or metrics. If a job posting requires something your ledger does not support, the gap is stated plainly instead of filled in.

2. No fetched sources

Drafts (recruiter emails, interview prep packs) read your verified ledger and the text you paste. They never add a claim, a company fact, or a citation that is not already supported

by you. Nothing is fetched from the internet to fill out your profile.

3. No training on your data

No vendor AI model is used by default. Generation is deterministic from your verified ledger facts — nothing is sent to a third party. If you enable a free local model (Ollama) or Hugging Face free tier, it runs with no-training / zero-retention settings where supported, and every output is re-checked against your ledger before it reaches you. Your ledger text is never used to train any model.

4. Providers

HonestFit uses free AI models only. Hugging Face's free inference tier is enabled by default, running with no-training settings where supported. Every AI output is re-checked against your verified ledger before it reaches you, so no invented skills, titles, or metrics ever appear in your artifacts. You can switch to Ollama (local, runs on your hardware) or disable AI entirely through configuration. No paid vendor models are ever used.

5. Your control

  • Edit or remove any verified fact at any time in the Facts Ledger.
  • Export or delete your data; nothing is held hostage to a plan.

More on how we protect your data: Privacy Notice.