Evidence · DiggingBeagle record
Proofpoint: BluePhishing through agent blueprint consent
Primary or origin disclosure used to qualify the underlying mechanism, stated prerequisites and limits. Proofpoint PoC: attacker blueprint AzureADMultipleOrgs, client credential + requiredResourceAccess; victim consents; tenant receives blueprint principal with AgentIdentity.CreateAsManager and attacker can create local agent identities through original credentials. Up to 250 identities/blueprint. Sponsor ID and tenant ID needed. Application permissions need admin consent; delegated grants need additional user/OBO context; inherited delegated permissions sometimes invisible in standard inventory. Microsoft MC1506733 changes default blueprint creation in Nov; creation policy separate from consent policy.
- Source role
- primary disclosure
Evidence record
Primary or origin disclosure used to qualify the underlying mechanism, stated prerequisites and limits. Proofpoint PoC: attacker blueprint AzureADMultipleOrgs, client credential + requiredResourceAccess; victim consents; tenant receives blueprint principal with AgentIdentity.CreateAsManager and attacker can create local agent identities through original credentials. Up to 250 identities/blueprint. Sponsor ID and tenant ID needed. Application permissions need admin consent; delegated grants need additional user/OBO context; inherited delegated permissions sometimes invisible in standard inventory. Microsoft MC1506733 changes default blueprint creation in Nov; creation policy separate from consent policy.
Claim-level citations (3)
- supportsBluePhishing agent-blueprint consent can create attacker-managed identities: Proofpoint describes up to 250 child agent identities per consented blueprint, but the manager permission, application grants and delegated user context must still be distinguished before concluding that any data access occurred.
Proofpoint BluePhishing report, agent blueprint consent flow, AgentIdentity.CreateAsManager and permissions discussion
- supportsBluePhishing agent-blueprint consent can create attacker-managed identities: Proofpoint expressly reported a proof of concept, not a verified live BluePhishing intrusion or stolen customer record.
Technical demonstration: multitenant agent blueprints, manager identity, consent requirements and no observed campaign
- supportsBluePhishing agent-blueprint consent can create attacker-managed identities: Proofpoint demonstrated a consent route in which an attacker-controlled multitenant Microsoft Entra agent blueprint can lead a consenting tenant to instantiate local agent identities under an attacker-retained manager credential.
Technical demonstration: multitenant agent blueprints, manager identity, consent requirements and no observed campaign