AI CMO security built around your data and your budget
An AI CMO touches two sensitive things, your company data and your marketing budget. VirtualCmo reads data through read-only connections, encrypts it, never uses it to train shared models, and only changes spend inside guardrails you set.
Encryption in transit and at rest
All traffic uses TLS. Databases, backups and connection tokens are encrypted at rest. Tokens for your ad, analytics and CRM accounts are stored encrypted and never shown in the interface after you connect.
Read-only scopes by default
Connections request read access. The only write permission is changing budgets in ad accounts, and only if you grant it while connecting and set guardrails for it. Disconnect any source with one click.
Guardrails on every automatic action
Approve mode is the default. Auto-execute only runs inside rules you set: maximum shift per channel, monthly spend cap and approval thresholds. Total spend is never raised automatically.
Roles and permissions
Owner, Admin, Approver and Viewer roles control who connects data, who approves spend and who only reads reports. Enterprise adds custom roles and an IP allowlist.
Single sign-on on Enterprise
SAML, Google and Microsoft sign-in with SCIM provisioning, so access follows your identity provider and leavers lose access when you remove them there.
Data residency choice
Enterprise workspaces choose US or EU hosting for their data, with retention controls for synced metrics and reports.
No training of shared models
Your data is used to plan, recommend and report for your workspace only. It is not used to train models shared with other customers, and AI model providers we use process it under terms that forbid training on it.
Audit log and decision log
Every recommendation, approval, automatic change and sign-in event is recorded. Scale and Enterprise expose the full audit log to Owners and Admins.
Security by plan
Every plan gets encryption, read-only connections, guardrails and the decision log. Larger organizations need identity and residency controls, which Enterprise adds. Details on pricing and integrations.
| Control | Growth | Scale | Enterprise |
|---|---|---|---|
| Encryption, read-only scopes, guardrails | Yes | Yes | Yes |
| Decision log | Yes | Yes | Yes |
| Roles and audit log | Owner and members | Owner, Admin, Approver, Viewer | Custom roles |
| SSO and SCIM | No | No | Yes |
| IP allowlist | No | No | Yes |
| Data residency US or EU | No | No | Yes |
| DPA | On request | On request | Download in settings |
| Uptime SLA | No | No | 99.9 percent |
Reporting a security issue
If you believe you have found a vulnerability, write to [email protected] with the steps to reproduce it. Please do not access other customers' data or degrade the service while testing. We reply to every report and fix confirmed issues as a priority.
Frequently asked questions
Can VirtualCmo spend our money without approval?
Only inside guardrails you set, and only if you grant budget permissions when connecting an ad account. By default every budget change waits for an Approver.
Do you sign a data processing agreement?
Yes. Write to [email protected] from your account email and we send our DPA for signature. Enterprise can download it from the workspace settings.
Is our data used to train AI models?
No. It is used only for your workspace, and our AI model providers process it under terms that do not allow training on customer data.
How do we delete our data?
Disconnect sources in the app to stop syncing, and delete the workspace in settings to remove it. You can also ask [email protected] to delete synced data from a specific source.
Where is our data hosted?
With a hosting provider in the US or the EU. Enterprise workspaces choose the region. See the privacy policy for the categories of processors.
See the plan your revenue target needs
Enter your website, revenue target and monthly budget. The required pipeline and three recommendations are ready in about a minute.