CVE-2026-91445: unauthenticated data access (non-AI)

CISO Take

This is an Oracle Siebel CRM End User flaw in the Communications Panel/Dashboard component that lets an unauthenticated attacker with HTTP network access read all accessible CRM data and modify some of it. It was fixed in the October 2026 Oracle Critical Patch Update. It has no meaningful AI/ML nexus: the 'ml_ui' category appears to be a misclassification, and no tracked AI package is affected. Exploitation is trivial (CVSS 8.2, AV:N/AC:L/PR:N/UI:N), but there is no EPSS data, no KEV listing, no public exploit and no Nuclei template at publication. Teams running Siebel CRM 17.0-26.9 should apply the October 2026 CPU promptly and avoid exposing the Siebel web tier directly to the internet. Treat it as a standard enterprise patching item, not an AI security priority.

Sources:NVDoracle.com

What is the risk?

Technical severity is high (CVSS 3.1 8.2): network-reachable, low complexity, no privileges, no user interaction, high confidentiality and low integrity impact. Exploitation likelihood is currently unknown (no EPSS, not in KEV, no public exploit or scanner template). AI/ML-specific risk is negligible because the product is a CRM application, not an AI/ML component. Exposure depends on whether the Siebel web tier is internet-facing.

How does the attack unfold?

Initial Access

Unauthenticated attacker sends crafted HTTP requests to an exposed Siebel CRM Communications Panel/Dashboard endpoint.

Impact

Attacker reads all End User accessible CRM data and makes limited unauthorized inserts, updates or deletes.

Initial Access

Unauthenticated attacker sends crafted HTTP requests to an exposed Siebel CRM Communications Panel/Dashboard endpoint.

Impact

Attacker reads all End User accessible CRM data and makes limited unauthorized inserts, updates or deletes.

How severe is it?

CVSS 3.1

8.2 / 10

EPSS

0.2%

chance of exploitation in 30 days

Higher than 16% of all CVEs

Source: EPSS v3 — FIRST.org

Exploitation Status

No known exploitation

Sophistication

Trivial

What is the attack surface?

AVNetwork

ACLow

PRNone

UINone

SUnchanged

CHigh

ILow

ANone

What should I do?

1 step

Apply the Oracle Critical Patch Update of October 2026 (https://www.oracle.com/security-alerts/cpuoct2026.html) to all affected Siebel CRM instances. Until patched, restrict network access to the Siebel web tier through VPN or IP allowlisting, and place a WAF in front of it. Monitor HTTP logs for unauthenticated requests to Communications Panel/Dashboard endpoints and for unusual bulk data reads or unexpected record changes. Review any downstream AI/ML or analytics pipelines that ingest Siebel data for possible tampering.

What does CISA's SSVC say?

DecisionTrack

Exploitationnone

AutomatableYes

Technical Impactpartial

Source: CISA Vulnrichment (SSVC v2.0). Decision based on the CISA Coordinator decision tree.

How is it classified?

Auth BypassData LeakageFramework

Which compliance frameworks are affected?

This CVE is relevant to:

NIST AI RMF

MANAGE 2.3-Responses to identified AI risks and unexpected events

Frequently Asked Questions

What is CVE-2026-91445?

This is an Oracle Siebel CRM End User flaw in the Communications Panel/Dashboard component that lets an unauthenticated attacker with HTTP network access read all accessible CRM data and modify some of it. It was fixed in the October 2026 Oracle Critical Patch Update. It has no meaningful AI/ML nexus: the 'ml_ui' category appears to be a misclassification, and no tracked AI package is affected. Exploitation is trivial (CVSS 8.2, AV:N/AC:L/PR:N/UI:N), but there is no EPSS data, no KEV listing, no public exploit and no Nuclei template at publication. Teams running Siebel CRM 17.0-26.9 should apply the October 2026 CPU promptly and avoid exposing the Siebel web tier directly to the internet. Treat it as a standard enterprise patching item, not an AI security priority.

Is CVE-2026-91445 actively exploited?

No confirmed active exploitation of CVE-2026-91445 has been reported, but organizations should still patch proactively.

How to fix CVE-2026-91445?

Apply the Oracle Critical Patch Update of October 2026 (https://www.oracle.com/security-alerts/cpuoct2026.html) to all affected Siebel CRM instances. Until patched, restrict network access to the Siebel web tier through VPN or IP allowlisting, and place a WAF in front of it. Monitor HTTP logs for unauthenticated requests to Communications Panel/Dashboard endpoints and for unusual bulk data reads or unexpected record changes. Review any downstream AI/ML or analytics pipelines that ingest Siebel data for possible tampering.

What systems are affected by CVE-2026-91445?

This vulnerability affects the following AI/ML architecture patterns: Enterprise CRM web applications, Data pipelines ingesting CRM data.

What is the CVSS score for CVE-2026-91445?

CVE-2026-91445 has a CVSS v3.1 base score of 8.2 (HIGH). The EPSS exploitation probability is 0.24%.

What is the AI security impact?

Affected AI Architectures

Enterprise CRM web applicationsData pipelines ingesting CRM data

Compliance Controls Affected

NIST AI RMF: MANAGE 2.3

What are the technical details?

Original Advisory

Vulnerability in the Siebel CRM End User product of Oracle Siebel CRM (component: Communications Panel/Dashboard). Supported versions that are affected are 17.0-26.9. Easily exploitable vulnerability allows unauthenticated attacker with network access via HTTP to compromise Siebel CRM End User. Successful attacks of this vulnerability can result in unauthorized access to critical data or complete access to all Siebel CRM End User accessible data as well as unauthorized update, insert or delete access to some of Siebel CRM End User accessible data. CVSS 3.1 Base Score 8.2 (Confidentiality and Integrity impacts). CVSS Vector: (CVSS:3.1/AV:N/AC:L/PR:N/UI:N/S:U/C:H/I:L/A:N).

Exploitation Scenario

An attacker scans for internet-exposed Siebel CRM instances and sends crafted unauthenticated HTTP requests to the Communications Panel/Dashboard component. They retrieve customer records and communication data and make limited modifications. If the organization uses CRM data to train or ground ML models, tampered records could poison those datasets, and exfiltrated data could support targeted phishing.