Six core principles, with the ones that can be enforced in code enforced there, not just written in a document. Every AI feature ships with guardrails, not just guidelines.
Our personal-details scrubber uses an AI step built to catch names from every cultural background, not just common English names, because a scrubber that misses unfamiliar names fails the people who most need privacy. It is automated and can miss a name. No user is disadvantaged by who they are.
Your vault is zero-knowledge: documents, roadmaps, and notes are encrypted on your device with a key we do not hold. Our AI never sees your vault, and the AI processing we do use runs only in the United States.
Our policies, data practices, and AI limitations are published in plain language. No hidden agendas, no black boxes.
Guardrails are built to keep roadmaps to government process information, not legal advice. Every step carries a confidence label. The AI tells you when it does not know something.
Designed for vulnerable populations first. Labeled for screen readers, with adjustable text, with a permanently free safety tier that needs no account, for the people who need help most.
An activity log of who did what in every business account, and accepted security risks recorded with an expiry date rather than waived quietly. A living commitment, not a marketing checkbox.
Free to download. Pro is optional, and safety mode never needs an account.
Available on iPhone, iPad and Android.