Teams should monitor for unexpected changes to the msDS-AllowedToActOnBehalfOfOtherIdentity attribute, especially on Domain Controllers and other high-value systems. Pair change detection with tight delegation controls, alerting on new computer objects, and immediate rollback of suspicious modifications. RBCD becomes dangerous when a controlled machine account is granted impersonation rights, so fast containment matters more than after-the-fact investigation.
How RBCD Abuse Turns an AD Permission Change into Lateral Movement
RBCD is not just another directory setting, it is a delegated impersonation path. Once a computer object can act on behalf of another account, the attacker can often turn that trust into access to services that would otherwise reject the original principal. That is why detection has to focus on the delegation relationship itself, not just on failed logons or obvious account takeover symptoms.
The most important practical question is whether the new delegation target makes sense for the system that received it. A change on a domain controller, file server, admin jump host, or other tier-zero system deserves immediate scrutiny because those systems can convert a single suspicious modification into broad access and privilege escalation.
Teams should also treat the creation of new computer objects as part of the same story. In many environments, RBCD abuse depends on an attacker first introducing or hijacking a machine account, then binding that account to a valuable target through the delegation attribute. Monitoring needs to join object creation, attribute change, and authentication activity into one chain of evidence.
What to Watch for in Detection and Triage
Change monitoring is the best early signal, but it has to be specific. Look for unexpected writes to Kerberos delegation security settings, especially when the target is a high-value host or the writer is not a known automation account. If the modification was not part of a planned delegation workflow, assume it is suspicious until proven otherwise.
Alerting should be tuned to the objects that matter most. A change on a low-value workstation may be a configuration mistake, but the same pattern on a domain controller, certificate server, or privileged admin system can be an active preparation step for lateral movement. That distinction should drive triage priority, not just the raw number of attribute edits.
Correlate the delegation change with Kerberos service ticket behavior, recent computer account creation, and any unusual use of privileged accounts. Attackers often try to blend the abuse into normal admin activity, so a single log source is rarely enough. The goal is to see the sequence: introduce or capture a machine identity, grant it impersonation rights, then use that trust to reach the next system.
Containment and Recovery When RBCD Is Suspected
The fastest containment step is to remove the suspicious delegation right and return the target object to its approved state. In parallel, isolate the source machine account or host that was used to stage the change, because the account itself may remain a launch point for more Kerberos abuse even after the attribute is rolled back.
Containment should not stop at the single modified attribute. If the attacker created a new computer object, reused an existing one, or obtained credentials that can still authenticate, those paths need to be closed before normal operations resume. Otherwise the same delegation pattern can be rebuilt quickly.
For longer-term hardening, map the event to broader delegation and privilege controls in the Active Directory and Entra ID hardening guide and the NHI lifecycle management guide. The practical lesson is that delegation abuse becomes much harder when machine objects, service identities, and privileged paths are tightly governed throughout their lifecycle rather than left to ad hoc administration.
Risk and Threat Considerations
RBCD abuse is dangerous because it converts a single directory permission into a reusable impersonation path. Once an attacker controls the machine account tied to that trust, they can often move laterally without needing a fresh password prompt or obvious interactive login, which makes the activity easy to miss if monitoring is too user-centric.
Failure mechanism: A malicious or compromised computer object is granted the ability to act on behalf of another identity, then requests service tickets that impersonate a more privileged principal. That breaks the normal boundary between a machine account and the services it should never be able to represent.
Impact: The attacker gains a low-friction route into higher-value systems, often with enough reach to stage lateral movement, access sensitive services, or pivot toward privileged administration paths before defenders notice the change.
Standards & Framework Alignment
This section maps relevant standards and security frameworks to the operational risks and controls described in this guidance.
MITRE ATT&CK addresses the attack and risk surface, while NIST CSF 2.0 and NIST SP 800-53 Rev 5 set the governance and control requirements practitioners need to meet.
| Framework | Control / Reference | Relevance |
|---|---|---|
| MITRE ATT&CK | T1134.001 — Access Token Manipulation: Token Impersonation/Theft | RBCD abuse relies on impersonation-style lateral movement and delegated authority. |
| Recommendation — Map suspicious impersonation paths to ATT&CK and hunt for adjacent lateral-movement activity. | ||
| NIST CSF 2.0 | DE.CM-01 — Networks and systems are monitored to detect cybersecurity events | Detecting RBCD abuse depends on monitoring directory changes and related authentication activity. |
| Recommendation — Monitor directory delegation changes and alert on anomalous computer-object activity. | ||
| NIST SP 800-53 Rev 5 | AC-6 — Least Privilege | RBCD abuse is a privilege expansion problem that least privilege should constrain. |
| AU-12 — Audit Generation | Change detection needs auditable records for delegation writes and related object creation. | |
| CM-5 — Access Restrictions for Change | Controlling who can modify delegation settings directly limits RBCD abuse paths. | |
| Recommendation — Restrict delegation rights to the minimum set of approved systems and accounts. Generate and retain logs for delegation attribute changes and computer-object creation. Limit who can change delegation attributes and review those changes promptly. | ||
Practitioner Guidance
What to prioritise: Put the highest response priority on delegation changes touching privileged targets, not on the loudest log volume. A small, precise modification to msDS-AllowedToActOnBehalfOfOtherIdentity can matter far more than a burst of routine authentication noise.
What to verify: Confirm who created the computer object, why the delegation was granted, and whether the target aligns with an approved delegation design. If you cannot establish a clear business owner and change record quickly, treat the change as hostile until the environment is proven clean.
Decision rule: If the suspicious delegation can authenticate to any system that supports administrative or tier-zero operations, contain first and investigate second. With RBCD, delaying rollback usually gives the attacker more time to use the trust relationship than the defender gains from extended observation.
Practitioner takeaway: The key containment decision is whether the delegation path itself is trustworthy, because once that trust is abused, the attacker can keep using it until the right is revoked and any associated machine account exposure is closed.
Related resources from NHI Mgmt Group
Deepen Your Knowledge
Free weekly newsletter
Subscribe to the NHI & AI Identity Journal
The latest on NHI and Agentic AI security – articles, research, breaches, news and events every week.
Bonus 33% off our NHI Course when you subscribe.
Reviewed and updated by the NHIMG editorial team on November 4, 2026.
NHI Mgmt Group — the #1 independent authority on Non-Human Identity, IAM, and Agentic AI security. nhimg.org