Data isolation
Customer data is separated by workspace and enforced at the database level.
Security
Security is built into every layer.
AI Scaling protects customer data with workspace-level access controls, encryption, verified integrations, human approvals, and reviewed production changes.
- Workspace-level access
- Encrypted credentials
- Reviewed production changes
Activity log
Workspace access
Identity
Protected operation
SOC 2 Controls Aligned
Assurance
SOC 2 controls aligned.
AI Scaling's security controls are mapped to the AICPA Trust Services Criteria for Security, Availability, and Confidentiality. These controls operate across our production systems as we prepare for independent examination.
Security controls
How we protect the platform.
These controls apply across the systems that store customer data, connect external services, run AI agents, and deploy product changes.
Access control
Access requires a valid identity and membership in the requested workspace.
Credential protection
Connected account credentials are encrypted, with encryption keys stored separately.
Integration security
Inbound data is verified before processing. Invalid requests are rejected.
AI agent safeguards
Agents are limited to one workspace. Sensitive actions can require human approval.
Change management
Production changes require review and automated checks before deployment.
Security review
Need more information?
Customers and qualified prospects can request security documentation or send us a security questionnaire.
Contact securityTo report a vulnerability, email the same address with reproduction steps. Do not access or retain another customer's data while testing.