---
title: "[AUTO] ServiceNow AI Platform: three unauthenticated CVSS 10.0 flaws"
pubDatetime: 2026-08-29T02:14:00.000Z
description: "Three maximum-severity vulnerabilities in ServiceNow AI Platform allow unauthenticated attackers to execute code and extract data via SQL injection."
tags: [ai-security, vulnerability, servicenow, 2026, 2026-q3, 2026-08, AUTO]
---
ServiceNow patched [three CVSS 10.0 vulnerabilities](https://support.servicenow.com/kb?id=kb_article_view&sysparm_article=KB3152242) in its AI Platform on August 27, 2026, any of which allows unauthenticated attackers to read backend data or run code. CVE-2026-18885 permits arbitrary code execution via the GraphQL Composite Data API; CVE-2026-18886 allows privilege escalation through the image upload processor; and CVE-2026-74820 exposes SQL injection in ORDER BY clauses. All three bypass authentication.

This combination is not unlucky. An attacker needs only one vector to establish a foothold; the others offer lateral movement and data extraction once inside. For ServiceNow's scale, that's a broad attack surface.

ServiceNow's claim of no current exploitation is carefully worded. Vendors typically mean they haven't seen confirmed attacks in their monitored environments, not that attacks haven't begun. This is the second major incident for the AI Platform in weeks, following [a critical flaw](https://thehackernews.com/2026/07/critical-servicenow-ai-platform-flaw.html) in July. The pattern suggests the platform needs harder security scrutiny before upgrading.

---

*Sources: [August 2026 CVE Advisory Notification (KB3152242)](https://support.servicenow.com/kb?id=kb_article_view&sysparm_article=KB3152242)*

*Coverage: [Three CVSS 10.0 ServiceNow Flaws Could Let Unauthenticated Attackers Execute Code and SQL](https://thehackernews.com/2026/08/three-cvss-100-servicenow-flaws-could.html)*

*Related on this blog: [[AUTO] AI-Generated Patches Fail at Scale](/posts/auto-ai-patches-fail)  •  [[AUTO] Langflow RCE Added to KEV After a Month of Silence](/posts/auto-langflow-rce-timing)  •  [[AUTO] When your AI agent decides unauthorized access is a reasonable tactic](/posts/auto-agent-gym-exploit)*