Policy Details and Scope
Define the identity, messaging, and deployment scope of a guardrail policy. These are the sections that decide what a policy is called, what users see when it blocks, and where it applies.
This page covers part of the Create Guardrail Policies form.
Provide Guardrail Policy Details (Mandatory)
This section defines identifying metadata and user-facing enforcement messages.
Name the policy
Enter a policy name and description.
Set the severity
Select the severity level to classify the importance of violations generated by the guardrail policy.
Define the blocked message
Define the blocked message shown when a request is denied.
Reuse the message for responses (optional)
Enable the option to apply the same blocked message to responses, if consistent messaging is required across requests and responses.

Scope
Configure which servers the guardrail should be applied to and specify whether it applies to requests, responses, or both. The coverage controls differ by product.
Akto Atlas: Agentic Assets
Choose which assets the guardrail is deployed across:
Apply to all (Recommended): Applies the guardrail to all assets. The panel shows the current counts, for example
46 Agents,95 MCP Servers, and22 LLMs, as links you can select to view the full list.Select Agentic Assets: Target specific assets using a condition builder.
To build a condition:
Choose an asset type
Under Where my, choose an asset type such as Agents, MCP Servers, or LLMs.
Select the values
Use the are field to search and select the matching values.
Layer or reset conditions
Select Add condition to layer additional conditions, or Clear all to reset the configuration.
Include or exclude specific assets
In both Akto Atlas and Akto Argus, each condition can either include or exclude its matches, so you can apply a policy broadly and carve out exceptions.
Include or exclude specific agents, MCP servers, LLMs, devices, and users.
Combine include and exclude conditions in the same policy, for example apply to all agents except a trusted internal agent.
Akto Atlas: Device Tags & Users
Choose which users and devices the guardrail applies to:
Apply to all (Recommended): Applies the guardrail to all users in your organisation, for example
136 Users.Select Device Tags & Users: Choose specific device tags or users using the same condition builder.
To build a condition:
Choose an attribute
Under Where my, choose an attribute such as Groups, a specific user, or a specific device.
Select the values
Use the are field to search and select the matching values. To target users by email address, select the user's email as the value.
Layer or reset conditions
Select Add condition to layer additional conditions, or Clear all to reset the configuration.
The number of matching users updates as you build the condition, for example Applies to 1 user with the current selection.

Akto Argus: Coverage
Akto Argus shows a single Coverage section instead of separate Agentic Assets and Device Tags & Users sections:
Apply to all (Recommended): Applies the guardrail to all discovered assets. The panel shows the current counts, for example
60 Agents & 31 MCP Servers, as links you can select to view the full list.Advance Configuration: Build a condition to target specific assets.
To build a condition:
Choose an asset type
Under Where my, choose an asset type such as AI Agents or MCP Servers.
Select the values
Use the are field to search and select the matching values.
Layer or reset conditions
Select Add condition to layer additional conditions, or Clear all to reset the configuration.

Akto Atlas: Work & Personal Accounts (Beta)
Apply the policy based on the account used to sign in to AI tools:
All accounts (Recommended): Applies to both work and personal accounts.
Personal accounts only: Applies to personal accounts only. Work accounts are excluded.
Use Personal accounts only to enforce stricter guardrails on usage outside your organisation-managed identities without affecting work accounts.
Rule Behaviour
Choose how Akto responds when a guardrail condition is triggered:
Block: Stops the request or response when the condition is met.
Alert: Generates an alert for review without blocking content.
Human Approval: Blocks the triggering attempt and routes it to your designated approvers through the Needs Approval tab in Guardrail Activity, where they approve or deny it. See Manage Human Approval Requests for the full workflow.
Application Settings
Specify whether the guardrail should be applied to responses and/or requests:
Apply guardrail to requests: Evaluates user inputs before processing.
Apply guardrail to responses: Evaluates model outputs before delivery.
You can enable either option independently or both together based on enforcement requirements.
Last updated