Threat Summary
CVE-2026-6875 is a critical remote code execution (RCE) vulnerability in the ServiceNow AI Platform. The flaw allows unauthenticated attackers to escape the sandbox and execute arbitrary code with elevated privileges on affected ServiceNow instances. Attackers may leverage this bug to gain persistent access, exfiltrate sensitive business information, manipulate workflows, or launch lateral attacks within the target environment.
The vulnerability affects both hosted and self-hosted versions of ServiceNow AI Platform prior to the following: Australia Patch 2, Yokohama Patch 12 Hot Fix 1b/Patch 13, Zurich Patch 7b/Patch 9, and Brazil EA/GA. ServiceNow patched all hosted instances as of July 13, 2026, but self-hosted clients must manually apply the fix. In-the-wild exploitation was reported on July 18, 2026.
A proof-of-concept was publicly released by Search Light Cyber, but the active exploit used by reported threat actors reaches the same code execution by a different route than their published proof-of-concept. The vulnerability is rooted in ServiceNow’s script sandboxing and access control logic. Attackers exploit misconfigured sandbox-enabled script includes or script evaluator APIs to escalate privileges and execute commands. The attack surface is significant due to ServiceNow’s widespread use, estimated at 85% of Fortune 500 companies. At the time of the bulletins writing, CISA/US-CERT has not yet issued a KEV listing.
The threat landscape remains active. Scanning and exploit attempts will continue as adversaries seek to abuse unpatched environments. Organizations with potential exposure since July 13, 2026 should also review logs and audit access patterns for suspicious activity.
Recommendations for CVE-2026-6875
- Immediately apply the ServiceNow security update or upgrade to patched platform versions.(Hosted (SaaS) environments are already patched; self-hosted customers must act urgently)
- Australia Patch 2 or newer
- Yokohama Patch 12 Hot Fix 1b/Patch 13+
- Zurich Patch 7b/Patch 9+
- Brazil EA/GA+)
- Audit platform logs for indications of exploitation (e.g., access to /assessment_thanks.do or suspicious unauthenticated script execution).
- Enable and enforce script sandboxing:
- Set glide.script.use.sandbox = true and audit for any unnecessary ‘sandbox enabled’ script includes
- Remove or restrict them where feasible.
- Harden client and backend script execution:
- Enable glide.ui.escape_all_script = true, glide.html.escape_script = true, and glide.html.sanitize_all_fields = true.
- Test changes in staging before broad deployment, as some UI functionality may be impacted.
- Review ‘Client callable’ script include usage and ensure no privilege escalation vectors exist. Limit the scope of scripts that can be executed by unauthenticated or low-privilege users.
- Continue to monitor ServiceNow advisories, threat intelligence feeds, and the CISA KEV catalog for ongoing updates or shifts in exploitation trends.
Temporary Workarounds
If patching cannot be applied immediately, use these interim mitigations:
- Enforce strict script sandboxing by setting glide.script.use.sandbox = true. IMPORTANT. Once set, this is a one-way operation and cannot be undone according to ServiceNow documentation. Validate impact before production use.
- Enable glide.ui.escape_all_script = true and glide.html.escape_script = true to limit HTML/JS injection vectors, and set glide.html.sanitize_all_fields = true for comprehensive sanitization.
- Audit for and disable any non-essential ‘sandbox enabled’ script includes. Remove the ‘sandbox enabled’ flag from custom includes unless absolutely necessary.
- Apply the full hardening settings from Security Center Baseline v7.0 and regularly audit your ServiceNow instance for configuration drift or unauthorized changes.
Note: These workarounds reduce risk but do not fully eliminate the vulnerability. Patching to a fixed release remains mandatory for comprehensive protection.
References
- https://www.bleepingcomputer.com/news/security/critical-servicenow-code-execution-flaw-now-exploited-in-attacks/
- https://slcyber.io/research-center/smashing-the-servicenow-sandbox-pre-authentication-rce/
- https://nvd.nist.gov/vuln/detail/CVE-2026-6875
- https://www.cyber.gc.ca/en/alerts-advisories/servicenow-security-advisory-av26-693
- https://www.servicenow.com/docs/r/platform-security/instance-security-hardening-settings/updated-hardening-settings-for-baseline-version-seven.html
