CVE-2026-18883
ServiceNow Email Response AI Agent was vulnerable to indirect prompt injection
ServiceNow · Now Assist for CSM · Email Response AI Agent
First documented disclosure · Review checkpoint
Summary
Untrusted content could influence the Email Response AI Agent.
Severity and exploitation evidence
High · 7.1/10 · CVSS 4.0. Severity source.
Exploitation not established: Reviewed public sources describe the weakness, a correction or a controlled demonstration; malicious exploitation against users is not established.
Technical severity is not a measure of observed attacks or of risk to your deployment. Attempts, demonstrations, and successful compromise retain the qualifications in the source.
How the weakness works
ServiceNow identifies indirect prompt injection; the public CNA does not detail the payload or downstream action.
Conditions required
The affected Email Response AI Agent processes attacker-influenced content; further prerequisites are not specified.
Potential impact
Agent behavior may be redirected; specific data or action impacts are not described in the public record.
Affected versions and configurations
CSM AI agent collection before v6.1.
Fix and mitigation status
Centrally deployed remediation; no platform upgrade or hot fix is required.
Practical next steps
Use the centrally remediated agent collection; no version-specific platform backport is prescribed.
Operating lesson
Distinguish centrally updated agent behavior from platform patch requirements.
Evidence and limits
ServiceNow states it is not aware of malicious exploitation. The linked KB3235180 returns only a loading shell through direct public retrieval; the manufacturer’s full CNA record supplies the verified detail. CNA publication is 2026-11-19; this is distinct from the earliest reviewed public source and correction dates. SN assessments: CVSS 4.0 7.1.
Sources
- SN CVE authority record
Primary source ·