Workspace content stays customer-controlled
Vault treats chats, files, notes, transcripts, and generated documents as workspace data governed by team access.
Trust Center
Security review for private AI adoption.
Vault is built for teams that need AI to fit security, privacy, procurement, and operational review. This page summarizes how Vault approaches customer data, model access, auditability, and responsible disclosure.
- No training on workspace content
- Workspace permissions
- Reviewable AI work
Last updated: May 5, 2026
At a glance
The security story reviewers need first.
The goal is simple: centralize AI work in a governed workspace so teams can understand what data is used, who has access, which models are available, and where outputs remain.
No training on customer workspace content
Vault is designed so customer workspace content is not used to train third-party AI models.
Admins can govern team workspaces
Workspace membership, SSO support, permissions, and model availability help teams reduce unmanaged AI usage.
Work remains easier to inspect
Shared context, generated work, and audit-oriented history stay together for permitted teammates and reviewers.
Security posture
Controls buyers expect before AI reaches sensitive work.
Vault focuses on reducing unmanaged AI usage by giving teams one governed workspace for model access, shared knowledge, files, meetings, and generated work products.
Data handling
Workspace content is treated as customer-controlled data and is designed to stay inside managed team workspaces.
- Files, chats, notes, transcripts, and generated documents are handled as workspace data.
- Vault is designed so customer workspace content is not used to train third-party AI models.
- Sensitive workflows should be configured under the right agreement and internal policy.
Access controls
Admins can organize access around users, teams, workspaces, and model availability.
- Workspace-level permissions for shared knowledge and collaboration.
- Administrative controls for user and workspace management.
- Single sign-on support for business and enterprise rollouts where enabled.
Reviewability
Vault keeps AI work, context, and collaboration closer together so reviewers can understand how outputs were produced.
- Audit-oriented activity history for important workspace actions.
- Shared documents, chats, files, and folders remain visible to permitted users.
- Security and legal pages support procurement and vendor-review workflows.
AI provider governance
Teams can bring multiple AI models into one approved workspace instead of scattering sensitive prompts across unmanaged tools.
- Centralized model access across supported providers.
- Buyer-review context for model availability, data handling, and usage boundaries.
- Configuration and provider commitments should be confirmed in the applicable customer agreement.
Data flow
A clearer operating model for AI work.
Security teams usually need to understand where data enters, how context is controlled, which model is used, and where the output remains after generation.
- 1
User adds context
A permitted user starts a chat, uploads a file, captures notes, or opens shared workspace knowledge.
- 2
Vault applies workspace controls
Workspace membership, model access, and configured team boundaries shape what context can be used.
- 3
Approved model responds
The selected model returns an answer or draft inside Vault, where the result can be reviewed and reused.
- 4
Work stays reviewable
Chats, documents, files, and activity history remain in the workspace for permitted teammates and admins.
Buyer review
Questions Vault is designed to help answer.
These are common questions from security, privacy, compliance, and procurement teams evaluating AI workspaces.
- What customer data is processed, stored, or sent to model providers?
- Which AI models and providers are available to users?
- Who can access shared workspaces, uploaded files, transcripts, and generated documents?
- How does the product support audit history and administrative review?
- What privacy, retention, subprocessors, and contractual commitments apply?
- Who should security teams contact for vulnerability reports or procurement review?
Procurement
Documents and next steps.
Public pages can answer the first layer of review. Detailed security materials, contractual terms, and regulated-data commitments should be requested through the review process.
Responsible disclosure
Found a security issue?
Email hello@vaultsystems.ai with enough detail for us to investigate. Please avoid accessing, changing, or disclosing data that is not yours.