Microsoft is preparing a change to Microsoft Entra that will allow ordinary users to create agent blueprints by default, potentially expanding how organizations build and manage AI-driven identities. The rollout is scheduled to begin on November 4, 2026, and is expected to finish worldwide by late November.
For IT administrators, the update presents an important permissions decision. Organizations that want to restrict who can create agent blueprints must review the Users can create agent blueprints setting in the Microsoft Entra admin center and switch it to No before the rollout takes effect.
The change does not require action from every organization. However, administrators who want to retain tighter control over agent creation should review their tenant settings now rather than assume existing permissions will remain unchanged.
What Is Changing in Microsoft Entra?
Microsoft Entra is changing the default permission that governs who can create agent blueprints within an organization.
Starting November 4, default users who do not hold a built-in or custom Microsoft Entra role will be able to create new agent blueprints unless an administrator disables the capability. The change will roll out globally in early November, with completion expected by late November 2026.
The key details are:
- Rollout begins: November 4, 2026.
- Expected completion: Late November 2026.
- Affected service: Microsoft Entra.
- Primary setting: Users can create agent blueprints.
- Required action: Administrators must change the setting to No if they want to restrict default users from creating new blueprints.
The setting can be configured before the rollout begins. However, Microsoft says that changing it now will not immediately alter the tenant’s behavior; the updated preference takes effect when the rollout starts.
This gives organizations a window to review their policies and prepare before the new default becomes active.
Why This Change Matters for IT Administrators
AI agents are becoming part of enterprise technology environments, where they may need identities and permissions to interact with applications, services, and organizational resources.
As organizations expand their use of agents, controlling who can create them becomes an increasingly important part of identity governance. A change to the default creation permissions can affect how IT teams oversee the development of these resources, even when the change does not automatically grant access to every connected system.
Microsoft’s update makes agent blueprint creation available to a broader group of users by default. That may suit organizations encouraging experimentation and employee-led AI development, but it may not align with environments where agent creation is restricted to designated teams.
For administrators, the decision comes down to balancing accessibility with governance.
Organizations should consider whether ordinary users need the ability to create agent blueprints, whether internal policies require centralized approval, and whether their existing identity management procedures adequately cover AI-related resources.
The important distinction is that permission to create a blueprint is not the same as unrestricted access to organizational data or resources. Administrators should evaluate the specific permissions and access requirements associated with their agents rather than treating blueprint creation as a complete measure of agent security.
How to Restrict Agent Blueprint Creation
Organizations that do not want default users creating new agent blueprints should update the relevant setting in the Microsoft Entra admin center.
Step-by-step instructions
- Open the Microsoft Entra admin center and sign in with an appropriate administrative account.
- Navigate to Entra ID > Users > User settings.
- Locate Default user role permissions.
- Find Users can create agent blueprints.
- Change the toggle to No and save the change.
Administrators should verify that the setting reflects their organization’s intended policy before the rollout begins.
Microsoft notes that making this change in advance will not cause an immediate change in behavior. Instead, the configured preference will take effect when the rollout starts in early November.
For the official guidance, consult Microsoft’s documentation on managing who can create agent blueprints.
Who can still create agents when the setting is disabled?
Disabling the setting does not prevent every authorized person from creating agents. Supported administrators and users assigned the Agent ID Developer role can continue to create them.
This distinction matters for organizations that want to limit creation to approved personnel without removing the capability from their development teams.
Rather than blocking all agent creation, administrators can use the setting to preserve a more controlled model in which designated roles retain the necessary permissions.
What Changes—and What Stays the Same?
Understanding the boundaries of this update is essential to avoid unnecessary disruption or incorrect assumptions about its security impact.
According to Microsoft’s announcement, the change affects the default user’s ability to create new agent blueprints. It does not:
- Modify or delete existing agent blueprints.
- Change who can create agent identities from existing blueprints.
- Block an application or an agent from creating identities.
- Override licensing, Azure subscription, or other creation requirements imposed by platforms such as Microsoft Foundry or Security Copilot.
These exclusions are significant because the update addresses a specific permission rather than replacing the organization’s entire agent management model.
For example, an organization that disables default-user blueprint creation does not thereby remove existing blueprints or automatically prevent identities from being created through every other supported mechanism. Administrators should review those separate pathways if their goal is to control agent deployment more broadly.
Likewise, organizations using Microsoft Foundry or Security Copilot must continue to account for the requirements of those platforms. Changing the Entra setting does not remove platform-specific prerequisites.
Do Not Confuse Agent Blueprints With Application Registration
Microsoft has also highlighted an important distinction between Users can create agent blueprints and Users can register applications.
These are separate settings and should be reviewed independently.
An organization may have an established policy for application registration that does not automatically reflect its preferred policy for agent blueprint creation. Assuming one setting governs both activities could leave administrators with an incomplete picture of their tenant’s configuration.
As part of the November rollout preparation, IT teams should review both settings and confirm that each aligns with their internal security policies.
This is particularly relevant for organizations that already restrict application registration to approved users but have not yet established an equivalent approach for agent creation.
What Organizations Should Do Before November 2026
The appropriate response depends on how an organization wants to manage agent development.
Microsoft’s guidance makes clear that no action is necessary if default users should be allowed to create agent blueprints. Organizations that want to restrict the capability, however, should configure the setting before the rollout takes effect.
A practical preparation checklist includes:
- Review the current policy: Determine which users should be permitted to create agent blueprints.
- Inspect the Entra setting: Check the current value of Users can create agent blueprints.
- Apply the intended control: Set the toggle to No if default users should not have this capability.
- Confirm authorized roles: Ensure supported administrators and users assigned the Agent ID Developer role retain the access required for their responsibilities.
- Review related permissions: Assess application registration and other identity creation pathways separately.
- Validate the configuration: Confirm that the saved setting matches the organization’s intended policy and account for the announced rollout timing.
Organizations with formal change-management processes may also want to document the decision and communicate it to identity administrators, security teams, and AI development groups.
The aim is not necessarily to prevent employees from experimenting with AI. It is to make sure that the level of access matches the organization’s governance requirements.
Security and Compliance Implications
Microsoft has identified no specific compliance considerations for this change. Organizations should still assess its implications in the context of their own regulatory obligations, internal controls, and risk management policies.
A default permission change can be operationally significant even when it does not introduce a new compliance requirement. The relevant question is whether the new default aligns with the organization’s existing rules for creating and managing identities.
For highly regulated environments or organizations with centralized technology approval processes, administrators may prefer to limit blueprint creation to authorized roles. Other organizations may choose to permit broader access while relying on additional approval procedures and monitoring.
Neither approach is universally appropriate. The right decision depends on the organization’s risk tolerance, development model, and ability to govern the resources its users create.
It is also important to avoid overstating what this single setting accomplishes. Restricting blueprint creation does not, by itself, establish comprehensive control over agent identities, application permissions, or downstream resource access. Those areas require their own reviews.
What Happens Next?
The worldwide rollout is expected to progress from early November through late November 2026. Organizations should prepare for the new default while recognizing that the setting will take effect according to Microsoft’s rollout schedule.
For IT administrators, the immediate priority is straightforward: decide whether default users should be able to create agent blueprints and configure the relevant permission accordingly.
The broader issue is how organizations will govern AI agents as they become more common across enterprise environments. Decisions about who can create these resources, which roles retain that capability, and how identities are managed will increasingly form part of routine identity administration.
Microsoft Entra’s upcoming change offers organizations an opportunity to review those decisions before the new default takes effect. A brief permissions review now can help prevent a mismatch between the platform’s default behavior and the organization’s intended access policy.
