ServiceNow AI Platform vulnerability enables unauthenticated user impersonation
A critical vulnerability in ServiceNow's AI Platform allows unauthenticated attackers to impersonate authorized users and access their permissions. ServiceNow discovered the flaw through AppOmni research and patched it in October 2025, with no reported active exploitation.
What happened
A critical vulnerability (CVE-2025-12420) in ServiceNow's AI Platform infrastructure enables unauthenticated attackers to impersonate legitimate users and execute unauthorized operations. The flaw was discovered by AppOmni and disclosed to ServiceNow in October 2025.
The privilege escalation vulnerability affects the Assist AI Agents (sn_aia) and Virtual Agent API (sn_va_as_service) components. ServiceNow deployed security patches on October 30, 2025, to hosted instances and provided updates to partners and self-hosted customer deployments.
Reported impact
Upon successful impersonation, attackers gain access to all operations and permissions associated with the compromised user account, potentially leading to unauthorized data access, configuration changes, and lateral movement within enterprise environments.
Index analysis
The vulnerability has a CVSS score of 9.3, indicating critical severity. ServiceNow reports no evidence of active exploitation in the wild as of the disclosure date.
Governance controls
ServiceNow strongly recommends that customers immediately apply security updates or upgrade to patched versions, with minimum versions specified as 5.1.18 or 5.2.19 for Assist AI Agents, and 3.15.2 or 4.0.4 for Virtual Agent API.
Classification
Sources and evidence
This record was researched and written by the Index, based on the following original reporting.