Security & data
How we protect your data.
Good security is practice plus promise. Here is how we build to keep your data safe, what we access, what gets stored, and what we will never do with the access you give us.
Four commitments we make to every client.
Your data stays in your systems
Wherever possible, your information stays where it already lives. We connect to your tools; we don't move your records out of them.
We access only what the task needs
If we are automating invoices, we ask for invoice access, not your whole drive. Scope is agreed in writing before we connect anything.
Everything is documented first
Before a project starts you get a written plan: what we will access, how data flows, and what gets stored. You approve it before we begin.
A human approves every change
The AI drafts, sorts, and suggests. Anything that changes your records waits for someone on your team to say yes.
The details
How your data moves.
- What we access
- Only the systems named in your approved plan, with the narrowest access that gets the job done. If the scope needs to grow, we ask first.
- Where processing happens
- AI features run on established model providers through their business APIs. Your data goes to them only to complete the task at hand, and the API terms we build on do not allow it to be used for training their models.
- What gets stored
- The workspace keeps what it needs to function: documents in review, workflow runs, report data, and action logs. We don't keep shadow copies of your systems.
- If we part ways
- Access is revoked, credentials are destroyed, and we walk you through everything that was stored so you can keep it or remove it.
The AI under the hood
We don't train our own models. Glant builds on the major AI model providers through their business APIs, and we choose them for reliability and data-handling terms, not hype. Ask us which providers your workspace would use and we will tell you exactly, along with why.
The short list
What we never do.
- Sell or share your data. With anyone, for any reason.
- Train AI models on your business data.
- Keep access or credentials after an engagement ends.
- Change records in your systems without an approval step.
Who can do what, inside the workspace.
Per-user logins
Everyone on your team gets their own account. No shared passwords, no mystery sessions.
Permissions per tool
You decide who sees which tools and which data. The controller and the new hire don't need the same view.
Action logs
Every action the AI takes is recorded: what ran, when, and on whose approval. You can review any of it, any time.
Approvals built in
Steps that touch your records pause for a human. That is not a setting you have to find; it is how we build.
Ask us anything.
Security questions deserve direct answers from the people doing the work. Send us yours.
Or email us directly: hello@glant.ai