AI & Data Usage Policy
Effective date · 1 January 2025
All documentsLast updated · 12 November 2026
AI & Data Usage Policy
Effective date: 1 January 2025
Last updated: 12 November 2026
This policy explains what data Filary Intelligence uses, where it goes, and what it is not used for. It complements the AI Terms of Use and the Privacy Policy.
1. When data is used
Filary Intelligence only processes your data when you use it — for example, when you ask a question, open an AI summary or request a suggestion. If you are on the Free plan, or on Plus and do not use Filary Intelligence, your workspace data is not sent to an AI provider. You can disable Filary Intelligence, and disable its access to your workspace data, in your settings. When it is disabled, your workspace data is not sent to an AI provider.
2. What data may be used
To answer a request, Filary Intelligence may use the parts of your workspace relevant to it:
- clients (names, business names, contact details);
- invoices (line items, amounts, taxes, dates, status);
- payments and payouts (amounts, status, dates, fees);
- projects and work;
- your prompt and the conversation in that session.
Filary sends only what is needed for the request. Full card numbers and bank login credentials are never sent, because Filary does not have them. Unless you have disabled Filary Intelligence or disabled this access in your settings, Filary Intelligence can access the content of your workspace, including your files.
3. Who processes it
Requests are processed by Filary and by one or more of our default AI providers — xAI Corp. (Grok), Google LLC (Gemini) and OpenAI, LLC (ChatGPT) — acting as subprocessors (see /legal/subprocessors). Each provider processes the data to generate the answer. A provider may retain requests for a limited time for security and abuse monitoring, according to that provider's terms, then delete them. If you connect your own AI provider, that provider processes the data under its own terms, including its retention rules.
4. No training on your content
Filary does not use your workspace content or your clients' data to train general-purpose AI models. For the default providers we operate (xAI, Google and OpenAI), we use their programming interfaces and we do not opt in to any setting that would allow them to train general-purpose models on your content, where the provider offers that choice. We have not independently verified every provider's training practices, and those practices can change. If you connect your own AI provider, that provider's terms govern whether your content may be used for training; Filary does not control that use.
We may use:
- feedback you choose to give (for example rating an answer) and
- aggregated or de-identified usage metrics (for example how often a feature is used, or error rates)
to improve Filary Intelligence. These do not include your clients' identifiable information.
5. Storage of prompts and outputs
Your prompts and the outputs generated may be stored in your workspace so you can see your history, for as long as your account remains open, unless you delete them sooner. They are treated like the rest of your workspace content: covered by the Privacy Policy and, for your clients' data, by the Data Processing Agreement. When you delete your account, they are deleted with it.
6. Human control
Filary Intelligence does not make decisions with legal or financial effect on you or your clients. It does not send messages, change invoices or move money without your approval. See section 2 of the AI Terms of Use.
7. Your rights
You can access, correct or delete the personal information processed by Filary Intelligence as described in the Privacy Policy. Questions: security@getfilary.com.
8. Changes
If we change how data is used for Filary Intelligence in a material way (for example a new AI provider or a different training approach), we will update this page and the Subprocessors list and notify you by email before the change applies.
Related documents
A question about this document: security@getfilary.com
