CVE-2026-86857

Think of this like an office building where everyone has a keycard, but a flaw in the door locks lets any employee's card open rooms they were never supposed to enter. In this case, any logged-in user of the ServiceNow AI Platform could potentially view data meant for other departments or roles.

This matters to you if your organization runs ServiceNow AI Platform, whether self-hosted or through a partner-managed instance. Any user with a valid login — not just admins — could exploit this, so it affects you regardless of how restrictive your internal permission settings are.

ServiceNow has already shipped fixes across every active release branch, which signals the company treats this as serious enough to patch broadly and quickly. There's rising press attention tied to a known threat actor group, and while exploitation hasn't been confirmed yet, the combination of easy access (just needing a valid login) and wide patch rollout means you should treat this as urgent and patch without delay.

Identify your exact ServiceNow release family (Yokohama, Zurich, or Australia) and current patch/hotfix level via the System Diagnostics or Instance Information page. Target the fixed hotfix for your branch: Yokohama Patch 13 Hot Fix 5a; Zurich Patch 10 Hot Fix 3b, Patch 10 Hot Fix 4a W32, or Patch 11 Hot Fix 3; Australia Patch 2 Hot Fix 4b W32, Patch 4 Hot Fix 3, or Patch 5. For ServiceNow-hosted instances, the vendor has already deployed the fix — confirm via your HI/support portal rather than assuming. Self-hosted and partner-managed instances must apply the hotfix manually; test in a non-production instance first since hotfixes can include unrelated changes. There is no documented workaround since the flaw is in server-side authorization logic, not a configurable setting — restricting unnecessary user accounts and auditing role assignments can reduce blast radius while patching is scheduled. No CISA KEV listing or mandated remediation date exists yet. Monitor ServiceNow audit logs for anomalous cross-scope data access by non-privileged accounts, and watch for updated advisories given the rising press attention referencing the ShinyHunters actor group.

Weakness: CWE-862 (Missing Authorization) allowing an authenticated, low-privilege user to bypass access control checks via standard network requests, no user interaction required. Not currently listed in CISA KEV; no public exploit code or confirmed in-the-wild exploitation has been reported. EPSS scoring places exploitation likelihood in the lower percentile with a flat trend, consistent with no confirmed active campaign. Press attention is rising due to a roundup mentioning the ShinyHunters actor, but this reflects media/buzz rather than documented exploitation. Vendor has released hotfixes across all three active release families (Yokohama, Zurich, Australia), indicating broad platform impact and treated as warranting prompt remediation despite the absence of confirmed in-the-wild activity.