SANTA CLARA, Sept. 3 — ServiceNow has shipped patches for four security flaws in its AI Platform, three of which carry the vendor’s maximum severity rating and could let unauthenticated attackers execute arbitrary code and SQL under certain conditions. The company published its advisory on August 27, saying it had already deployed a security update to its hosted instances. Partners and self-hosted customers received the update as well, meaning organizations running their own instances must apply the fixes themselves.
ServiceNow is the CVE Numbering Authority for its products, so the ratings are its own. The three maximum-severity flaws share a CVSS vector describing a network-reachable attack of low complexity requiring no privileges and no user interaction, with high impact to confidentiality, integrity and availability both in the vulnerable component and in connected systems.
The company scored each at the top of the scale under CVSS version 4.0.
What the patches cover
A fourth flaw, tracked as CVE-2026-6876, allows an unauthenticated user to execute arbitrary code within the Now Platform. Its assigned vector, however, specifies low privileges required and records no impact to systems beyond the vulnerable component — distinguishing it from the three rated at maximum severity. ServiceNow listed affected versions in the August advisory.
The record for CVE-2026-18886 marks “Any version before Australia Patch 5” with a status of unknown, while the other three mark the same version as affected. All four records set a default product status of unaffected, and ServiceNow said in each case that it is not currently aware of exploitation.
Earlier sandbox escape
The August advisory follows CVE-2026-6875, a pre-authentication sandbox escape in the same platform that Searchlight Cyber reported to ServiceNow on April 1. ServiceNow published an advisory for that flaw on July 13. Days later, threat intelligence firm Defused said it was observing in-the-wild exploitation of CVE-2026-6875 — then issued a correction stating that the captured payload matched Searchlight Cyber’s published proof-of-concept exploit.
A ServiceNow spokesperson told The Hacker News: “ServiceNow is aware of a cybersecurity company’s recent publication regarding exploitation activity associated with a previously disclosed security vulnerability, identified as CVE-2026-6875.
Based on our investigation to date, we have not observed evidence that this activity is related to instances that ServiceNow hosts.” The spokesperson added that the company has “provided updates and patches designed to address this issue, and we encourage our self-hosted and ServiceNow-hosted customers to apply the relevant patches if they have not already done so.” ServiceNow scored the July sandbox escape at 9.5 under the same scoring version, with every metric identical to the newer flaws except attack complexity, set to high.
Adam Kues, a security researcher at Searchlight Cyber, wrote in July that ServiceNow was “enhancing instance security by severely restricting the type of code that can run in sandbox contexts.” Since April 15, NIST has enriched only vulnerabilities that appear in CISA’s Known Exploited Vulnerabilities catalog, affect federal government software, or are designated critical under an executive order. None of the four flaws appeared in the catalog as of August 28, leaving ServiceNow’s ratings as the only severity assessment on record.
No public exploit code for the three maximum-severity flaws had surfaced as of that date, and Searchlight Cyber had published no technical write-up for the August-disclosed flaws. For enterprises across India and Asia-Pacific running self-hosted ServiceNow deployments, the practical question is speed: the vendor has patched its own cloud, but on-premises customers carry the responsibility. The company says it will continue to work directly with customers who need assistance applying the patches.
