EUVD-2026-88208
GHSA-hwqr-vwwq-mq54
Attack ChainAIDerived
Hypothetical attack flow derived from CVE metadata – not a confirmed exploit chain
-
AccessAuthenticate as low-privilege GCP user
-
DeliveryAccess Application Integration task config
-
ExploitSelect internal-only task type
-
ExecutionIssue arbitrary internal RPCs under privileged identity
-
ImpactCompromise Google production-network resources
Vulnerability Assessment
| Weakness |
CWE
OWASP Top 10:2025
A01:2025 - Broken Access Control
ASVS 5.0
V8 - Authorization Authorization logic validation
MITRE ATT&CK
1 technique across 1 tactic: Privilege Escalation The technique list with ATT&CK links and the kill-chain view is part of the Basic plan and up. |
| Exploitation Conditions | Exploitation required an authenticated Google Cloud user with access to Google Cloud Application Integration and the ability to configure an integration task referencing an internal-only task type that should not have been available to tenants. Limiting factors: valid GCP authentication was required (PR:L, not unauthenticated), and the affected code path existed only in versions prior to the 2026-06-17 fix. As of the patch date the condition is fully remediated server-side by Google with no customer action possible or needed, so no current deployment is exploitable. |
| Impact | The CVSS 4.0 score of 9.4 (vendor) reflects a genuinely severe design flaw: an incorrect-authorization bug (CWE-863) that let an authenticated Google Cloud user (PR:L) invoke an internal-only task type to execute arbitrary internal RPCs from inside Google's own production network under a privileged identity - a true privilege-escalation/scope-crossing condition (SC/SI/SA:H). However, real-world residual risk is now effectively nil: this is a Google-operated managed SaaS, the vulnerability was remediated server-side on 2026-06-17, and Google explicitly states no customer action is needed. There is no EPSS score, no CISA KEV listing, and no public POC in the provided data - consistent with a vendor-discovered-and-fixed cloud issue rather than an in-the-wild threat. The high base score therefore describes the pre-patch potential impact, not an actionable ongoing risk for defenders. No conflicting signals in the vector itself, though the reporter is the vendor (GoogleCloud) so exploitation was likely identified internally. |
| Exploit ScenarioDerived | A user with legitimate but low-privilege access to Google Cloud Application Integration configures an integration task using an internal-only task type that should not have been exposed to tenants. Because authorization on the task configuration was not correctly enforced, the platform executes attacker-chosen internal RPCs from within Google's production network under a privileged service identity, crossing the tenant/production trust boundary. No public exploit code is identified, and the flaw was patched server-side before any customer-facing exposure remained. |
Get an e-mail when this CVE is exploited, patched or re-scored. Free account.
Alert meAnalysisAIv1 – Sep 28, 2026
Authenticated Google Cloud users holding permission to configure Application Integration tasks could abuse an incorrectly authorized, internal-only task type to trigger arbitrary internal RPCs that executed inside Google's production network under a privileged Google identity. The flaw (CWE-863) affected Application Integration on Google Cloud Platform in all versions prior to the 17 June 2026 fix, and the vendor's CVSS 4.0 vector of 9.4 reflects a genuine privilege-escalation and scope-crossing condition (SC/SI/SA:H) rather than simple data exposure. Google remediated the issue server-side on 17 June 2026 and states no customer action is required, meaning no current deployment remains exploitable; no public exploit code has been identified at time of analysis, and CISA KEV does not list this issue.
RemediationAI
Vendor-released patch: the issue was remediated server-side by Google on 17 June 2026 (bulletin gcp-2026-064), and because Application Integration is a fully managed service with no customer-deployed component, no upgrade, hotfix, or configuration change is required or possible on the customer side - Google explicitly states that no customer action is needed, and the vulnerable code path no longer exists in the hosted service. Organizations that want to verify or harden their posture should cross-check the two advisories at https://docs.cloud.google.com/support/bulletins#gcp-2026-064 and https://docs.cloud.google.com/application-integration/docs/security-bulletins#gcp-2026-064, and can as a preventive measure audit and tighten IAM bindings that grant Application Integration authoring or editing permissions (for example, roles that allow creating or modifying integrations and their tasks), since the flaw's prerequisite was the ability to configure a task referencing an internal-only type; the trade-off is that over-restriction slows legitimate integration development and may need to be balanced with break-glass access. Reviewing Cloud Audit Logs for Application Integration over the exposure window (before 17 June 2026) for unexpected task types, anomalous internal RPC patterns, or unusual identities is worthwhile as a detection and incident-scoping step, though Google's advisory indicates the condition is fully closed and there is no evidence of exploitation. No public exploit code has been identified at time of analysis, so there is no meaningful benefit to disabling Application Integration features or blocking network access as a compensating control.
Affected VersionsENISA
Affected Packages
Source: OSV (Open Source Vulnerabilities)
Affects your stack
Affects 1 package in euvd
Checking this CVE against your own stacks and servers is part of the Basic plan and up.
Technical ContextAI · show details
Application Integration is Google Cloud's fully managed integration platform-as-a-service, in which tenants compose 'integrations' from a catalog of tasks (triggers, data mapping, connectors, and callouts to Google and third-party services). Some task types in that catalog are internal-only: they are constructs used by Google's own services and are never intended to be selectable by tenant-authored configurations, because invoking them causes the backend to emit internal RPCs on Google's production network under a privileged Google service identity rather than the caller's identity. CVE-2026-19759 is an instance of CWE-863 (Incorrect Authorization): the authorization check that should have rejected tenant configurations referencing such an internal-only task type was missing or wrong, so an authenticated, low-privileged tenant user (PR:L in both the CVSS 4.0 and the assessed CVSS 3.1 vector) could cause code to run with a far more privileged identity and reach internal-only endpoints - the trust-boundary crossing that produces the SC/SI/SA:H scope-change metrics. The CPE string cpe:2.3:a:google_cloud:application_integration:*:*:*:*:*:*:*:* denotes the hosted service across all versions prior to the fix date rather than a customer-installable artifact; the vulnerable code path existed only in builds predating 17 June 2026 and was fully removed server-side. No authentication bypass was possible (a valid Google Cloud identity and Application Integration access were prerequisites), and there is no exploit chain published for the issue.
Know when Application Integration gets its next vulnerability
- Watch products and get an email when something new is published
- Check this CVE against your own stack and servers
- Exploit and PoC references, threat intelligence from 22 sources
No credit card · 7-day full trial
Severity by source
CVSS:4.0/AV:N/AC:L/AT:N/PR:L/UI:N/VC:H/VI:H/VA:H/SC:H/SI:H/SA:H/U:Clear
Authenticated GCP user (PR:L) over the network exploits broken authorization to cross into Google's production trust boundary (S:C) with full RPC execution under a privileged identity, yielding C/I/A:H.
Primary rating from Vendor (GoogleCloud).
CVSS VectorVendor: GoogleCloud
Lifecycle Timeline
2DescriptionCVE.org
An Incorrect Authorization vulnerability in the task configuration in Google Cloud Application Integration versions prior to 2026-06-17 on Google Cloud Platform allows an authenticated Google Cloud user to execute arbitrary internal RPCs from inside Google's production network under a privileged identity using an internal-only task type.
This vulnerability was patched on 17 June 2026, and no customer action is needed.
Executive ViewAI
Google Cloud Application Integration contained a critical privilege-escalation flaw (CVE-2026-19759, CVSS 9.4) that could have allowed users with task-configuration permissions to execute internal RPCs under a privileged Google identity within Google's production network. Google remediated the issue server-side on 17 June 2026 and states no customer action is required, so organizational exposure is closed; no public exploit identified at time of analysis. Leadership should care because the flaw crossed security boundaries at the cloud provider, but the vendor has eliminated the risk without customer intervention.
Restrict Application Integration task-configuration permissions to least privilege, enable audit logging and alerting for anomalous internal RPCs, and confirm server-side remediation with Google Cloud support. No customer patch is required per vendor; no vendor-released patch identified at time of analysis.
Threat Intelligence Overview
NVD
CVE.org
EPSS
Vulnrichment
How this decision was reached
The engine reads each signal below and applies a fixed list of rules from top to bottom. The first rule that matches sets the recommended action – like a statement, every line is shown.
What every action level means
SSVC AssessmentENISA
Priority Score
More in Application Integration
View allSandbox escape in the JavaScript Task feature of Google Cloud Application Integration allows an actor to execute arbitra
Broken access control in Google Cloud Application Integration's QueryEngineTask lets an unauthenticated external attacke
Privilege escalation in Google Cloud Application Integration (versions prior to 2026-08-27) allows an authenticated low-
Exploit-DB