All policies

AI Usage Policy

How Founder FingerprintsOS uses AI, what it does with your content, and where we deliberately stop.

Draft — pending legal reviewVersion 0.1.0Updated 8 August 2026

Draft — pending legal review.This document describes our intended practice but has not completed legal review, and is not yet in force. Please don't rely on it as a binding agreement until it is marked active with an effective date.

Our position on AI

Founder FingerprintsOS uses AI to help you think, not to think for you. The distinction matters: a tool that drafts an option for you to judge is useful, while a tool that quietly makes the judgement removes the accountability that running a company requires.

Every AI feature is something you invoke deliberately. We do not run background analysis over your ventures, and we do not generate decisions without being asked.

What happens to your content

When you use an AI feature, the content needed for that request is sent to a third-party model provider and the response is returned to you.

  • Only the content relevant to your request is sent, not your whole workspace.
  • We contract with providers on terms that prohibit retaining your content for model training.
  • Where a provider will not agree to those terms, we do not route your content to them.
  • We do not use your content to train models that serve other customers.

What AI here cannot do

AI output can be confidently wrong. It can invent figures, misread context, and produce reasoning that reads well but does not hold. Treat everything it produces as a draft to check.

Founder FingerprintsOS does not provide legal, financial, tax, or medical advice, and AI-generated text does not become advice because it sounds authoritative. For consequential decisions, verify against primary sources and consult a qualified professional.

Decisions we keep human

Some decisions are excluded from AI assistance because being wrong is materially harmful and the reasoning must be attributable to a person.

  • Terminating or disciplining an employee.
  • Final equity allocation or founder compensation.
  • Statements made to regulators, auditors, or courts.
  • Interpreting a binding contract you are about to sign.

Knowing when AI was involved

AI-generated content is labelled as such in the interface. Where an AI feature contributed to a stored artefact, that involvement is recorded with it, so a decision reviewed months later still shows how it was reached.

Your controls

AI features are opt-in per action, so declining to use them costs you nothing else. Where an organisation needs AI features disabled entirely, an administrator can request that and we will apply it at the workspace level.