Security & trust

Trust isn’t a feature.
It’s the product.

An agent that acts on your behalf has to be one you can hold accountable. Here’s how Foresigxt is built to earn that — and an honest account of where we are in the beta.

How your data is handled

Safe by
construction.

Glass-walled, so auditable.

Every step, tool call and source is logged and visible. If an agent touched your data, you can see exactly when, how and why.

Not your training data.

Your conversations, files and run data are never used to train models. They're yours; we just process them to do the job.

Least-privilege Skills.

A Role can only use the Skills you switch on. Turn off web access and it physically cannot reach the web. Nothing runs unsanctioned.

Encrypted, end to end of transit.

Data is encrypted in transit (TLS 1.2+) and at rest. Secrets and integration credentials are stored in an isolated vault.

Human-in-the-loop.

High-stakes actions can require explicit approval before an agent proceeds. You decide where the guardrails sit.

A full audit log.

Every run is recorded with its inputs, steps, cost and outcome — an exportable trail for review and compliance.

Where we are, honestly.

We’re a private beta with no commercial release yet. Some enterprise assurances are in flight rather than finished — we’d rather tell you that plainly than imply otherwise.

Encryption in transit & at rest Live
No-training-on-your-data policy Live
Per-run audit log Live
Role-level data controls Live
SOC 2 Type II For GA
SSO & SCIM For GA
Data-residency options For GA
Private / VPC deployment For GA

Found something? Report it to security@foresigxt.com — we read every report.

Questions on trust?

We’ll answer
straight.

Evaluating Foresigxt for a team with real requirements? Join the list and tell us — we’ll get you the detail and a beta seat.