Plans & OrganizationEnterprise Features
Model Governance & Access Control
Strategically manage which AI engines and providers are authorised for use across your engineering department.
Model Governance & Access Control
The Model Governance module provides administrators with absolute control over the AI landscape within their organization. You can precisely authorise or restrict specific AI engines and providers to align with corporate data policies and budgetary goals.
Administering Model Availability
- Access the Model Governance section within the Enterprise Portal.
- Utilise the selection matrix to enable or disable specific providers and individual models.
- Select "Authorize Changes" to propagate settings across the department.
Coming Soon !!!
Advanced Governance Filters
Engineers can only access models that cross-reference with your established filters:
| Strategic Filter | Governance Utility |
|---|---|
| Compliance Tier | Authorise only those models that adhere to specific data retention and privacy mandates. |
| Provider Geography | Restrict AI traffic to specific geographic regions to meet sovereignty requirements. |
| Model Generation | Standardize on specific families (e.g., Claude 3.5, GPT-4o, Gemini 2.0). |
| Authorized Providers | Whitelist entire providers like Anthropic, Google, or OpenAI. |
| Capability Matrix | Filter by model modalities (Code, Vision, High-Context, etc.). |
| Cost Efficiency | Compare and restrict models based on token expenditure tiers. |
Strategic Use Cases
- Privacy-First Environments: Automatically disable models that retain user prompts or operate outside authorized jurisdictions.
- Financial Oversight: Limit access to high-cost reasoning models for routine development tasks.
- Mission Specificity: Enable specialized code-centric engines while disabling general-purpose creative ones.
Critical Policy Notes
- Privilege Mandate: Only Administrators and Owners can calibrate model governance settings.
- Rapid Propagation: Governance updates are pushed to all developer environments within seconds.
- Centralized Enforcement: Organization-level restrictions always override individual user preferences.