Security and data handling.

Straterai builds AI systems on your company's data. Your team controls who can access it. Our default AI model, Claude, runs under zero data retention.

Default AI model
Anthropic API with zero data retention
Access
Controlled by your team
Encryption
TLS 1.2+ in transit, AES-256 at rest
Incidents
Notice within 72 hours
01

Access Control

We connect to the systems you already run. Your company controls our access levels.

02

Zero Data Retention

Our default AI models run through Anthropic's API under a zero data retention agreement. Anthropic does not store inputs or outputs and does not use your data for training.

03

Role-Based Access

Sign-in runs through your identity provider, with SSO, MFA and passkeys where it supports them. Each employee has scoped access to data and tools.

04

Least-Privilege Access

Your company grants us access with a read-only default. Our teams only read, access and utilize what is needed.

05

Data Ownership

Your data stays with your company, in your own environment wherever we can build there. Anything we hold for your project, we return or destroy on request.

06

Encryption

Data is encrypted in transit with TLS 1.2+ and at rest with AES-256. Every vendor that can access your data is listed on our sub-processors page.

AI model providers

  • ClaudeRecommended
  • OpenAI
  • Gemini
  • Microsoft Copilot

Cloud providers

  • Google CloudRecommended
  • Snowflake
  • Microsoft Azure
  • AWS

Build and control processes.

01

Architecture

Each company gets its own system.

Instance
Separate for each client. Your statement of work sets where it runs and whose account holds it.
Integrations
Read-only connections to your systems, granted by your team.
Writes
Only into the systems you name, when you ask.
02

Data residency

Your data stays in the region you choose.

Region
United States by default for US companies, or your existing cloud region.
Copies
We keep working notes, documents and databases for your project. On request, we return or destroy them and confirm it in writing.
Backups
Production databases are backed up, and recovery is tested.
03

Access and permissions

Permissions match the systems your team already uses.

Sign-in
Through your identity provider. Accounts are removed when a person leaves.
Permissions
Role-based and enforced at the database. Requests above a person's role are declined and logged.
Personal data
Employee and customer records are handled as personal data, one record at a time.
04

AI models

Each workflow sends the model only the data it needs.

Default model
Claude, from Anthropic. We also work with OpenAI, Gemini and Microsoft Copilot.
Anthropic API
Zero data retention, contracted with Anthropic and verified at the organization level.
Training
Prohibited under Anthropic's commercial terms.
Credentials
Each system uses credentials you approve.
05

Controls

Enforced in code and infrastructure, with evidence kept for audit.

Encryption
TLS 1.2 or higher in transit. AES-256 at rest. Full-disk encryption on every device with production access.
Our access
Single sign-on with multi-factor authentication on production systems.
Incidents
A documented incident response process. You are notified within 72 hours of a confirmed incident affecting your data, or sooner if your contract requires.
06

Due diligence

Your security and IT teams can review the system before it touches production data.

Inspection
We show your team how the system works, where data flows and what the model is allowed to do.
Evidence
Policies, access registers, the data inventory and recovery test records, available to your counsel under NDA.
Vendors
Listed on our sub-processors page and in your data processing addendum. We give 30 days' notice before adding one.
Our sub-processors

Compliance program.

In place

Security policies

Access control, data classification, acceptable use, backup and recovery, vendor management.

In place

Continuous monitoring

Through Vanta, across cloud accounts, source code and hosting.

In place

Network assessment

Completed May 2026. Repeated every year.

In place

Evidence registers

Access, devices, data inventory and recovery tests, kept in version control.

Scheduled

Penetration test

Independent test scheduled for 2026, then every year.

Security questions: security@straterai.com