Critical ServiceNow Flaw Allows User Impersonation; Urgent Patch Recommended

Critical ServiceNow Vulnerability Allows Unauthenticated User Impersonation and Privilege Escalation

A critical security vulnerability has been identified in ServiceNow’s AI Platform, designated as CVE-2025-12420, which enables unauthenticated attackers to impersonate legitimate users and escalate privileges within the system. This flaw poses significant risks, including unauthorized data access, configuration changes, and potential lateral movement within enterprise environments.

Discovery and Disclosure

The vulnerability was discovered by AppOmni, a SaaS security firm, and reported to ServiceNow in October 2025. Upon notification, ServiceNow promptly initiated remediation efforts to address the issue.

Technical Details

CVE-2025-12420 is a privilege escalation vulnerability within ServiceNow’s AI Platform infrastructure. It allows attackers without authentication credentials to assume the identity of authorized users. Once impersonation is successful, threat actors gain access to all operations and permissions associated with the compromised user account.

Potential Impact

The exploitation of this vulnerability could lead to:

– Unauthorized Data Access: Attackers may access sensitive information without proper authorization.

– Configuration Changes: Malicious actors could alter system configurations, potentially disrupting services.

– Lateral Movement: Gaining initial access could allow attackers to move laterally within the network, compromising additional systems.

ServiceNow’s Response

ServiceNow addressed the vulnerability on October 30, 2025, by deploying security patches to the majority of hosted instances. The company also provided updates to partners and self-hosted customer deployments. The vulnerability has been resolved in specific Store App versions released as part of the October 2025 security maintenance cycle.

Affected Applications and Patched Versions

The vulnerability impacts two critical ServiceNow applications:

– Assist AI Agents (`sn_aia`): Versions 5.1.18 or later, or 5.2.19 or later.

– Virtual Agent API (`sn_va_as_service`): Versions 3.15.2 or later, or 4.0.4 or later.

ServiceNow strongly recommends that customers immediately apply the appropriate security updates or upgrade to patched versions if deployment has not already occurred.

Recommendations for Organizations

Organizations operating both hosted and self-hosted ServiceNow environments should prioritize remediation of this vulnerability due to its critical nature and the potential for increased exploitation following public disclosure. Currently, ServiceNow reports no evidence of active exploitation in the wild. However, the window between public disclosure and widespread attack implementation is typically narrow, necessitating urgent action from security teams.

Conclusion

The discovery and prompt remediation of CVE-2025-12420 underscore the importance of proactive vulnerability management and collaboration between security researchers and vendors. Organizations relying on ServiceNow’s AI Platform components should review the complete security advisory and implement the recommended updates to safeguard their systems against potential exploitation.