PatchSiren cyber security CVE debrief
CVE-2026-5757 Ollama AI CVE debrief
CVE-2026-5757 is an unauthenticated remote information disclosure vulnerability in Ollama's model quantization engine. This vulnerability allows an attacker to read and exfiltrate the server's heap memory, potentially leading to sensitive data exposure, further compromise, and stealthy persistence. The vulnerability has a CVSS score of 7.5 and is classified as HIGH severity. The CVE was published on June 26, 2026, and last modified on June 29, 2026. The vulnerability affects Ollama versions up to 0.13.5.
- Vendor
- Ollama AI
- Product
- Ollama
- CVSS
- HIGH 7.5
- CISA KEV
- Not listed in stored evidence
- Original CVE published
- 2026-06-26
- Original CVE updated
- 2026-06-29
- Advisory published
- 2026-06-26
- Advisory updated
- 2026-06-29
Who should care
Organizations using Ollama versions up to 0.13.5 should prioritize patching this vulnerability to prevent potential sensitive data exposure and further compromise. Security teams should review their inventory of Ollama installations and ensure that all instances are updated to a patched version. Additionally, defenders should monitor for potential exploitation attempts and implement compensating controls to detect and prevent suspicious activity.
Technical summary
The vulnerability exists in Ollama's model quantization engine, which allows an attacker to read and exfiltrate the server's heap memory without authentication. The vulnerability has a CVSS vector of CVSS:3.1/AV:N/AC:L/PR:N/UI:N/S:U/C:H/I:N/A:N, indicating that it can be exploited over the network with low attack complexity and no privileges required. The vulnerability is classified as CWE-125, which is a weakness related to out-of-bounds read.
Defensive priority
High priority should be given to patching this vulnerability, as it can lead to sensitive data exposure and further compromise. Defenders should review their inventory of Ollama installations and ensure that all instances are updated to a patched version.
Recommended defensive actions
- Review and update Ollama installations to ensure that all instances are running a patched version.
- Monitor for potential exploitation attempts and implement compensating controls to detect and prevent suspicious activity.
- Implement additional security measures to protect sensitive data and prevent further compromise.
- Review and update incident response plans to ensure preparedness in case of a potential breach.
- Conduct a thorough risk assessment to identify potential vulnerabilities and prioritize remediation efforts.
Evidence notes
The CVE-2026-5757 vulnerability was identified in Ollama's model quantization engine, which allows an attacker to read and exfiltrate the server's heap memory without authentication. The vulnerability has a CVSS score of 7.5 and is classified as HIGH severity. The CVE was published on June 26, 2026, and last modified on June 29, 2026. The vulnerability affects Ollama versions up to 0.13.5.
Sources and references
Verified primary and authoritative sources
-
CVE-2026-5757 CVE Program record
Publisher, destination, and source semantics verified
URL: https://www.cve.org/CVERecord?id=CVE-2026-5757
CVE Program - Official CVE Program record with source-provided CVE metadata.
-
CVE-2026-5757 NVD vulnerability detail
Publisher, destination, and source semantics verified
URL: https://nvd.nist.gov/vuln/detail/CVE-2026-5757
NIST National Vulnerability Database - Official NIST NVD detail page and source-specific vulnerability assessment.
Supplemental references
-
Mitigation or vendor reference
Unverified legacy reference
URL: https://kb.cert.org/vuls/id/518910
[email protected] - Third Party Advisory, VDB Entry
-
Source reference
Unverified legacy reference
URL: https://ollama.com/
[email protected] - Product
-
Mitigation or vendor reference
Unverified legacy reference
URL: https://www.kb.cert.org/vuls/id/518910
af854a3a-2127-422b-91ae-364da2661108 - Third Party Advisory, VDB Entry
Methodology and review provenance
AI-assisted synthesis based on stored public vulnerability evidence. System validation, approval state, and publication status do not by themselves establish human review of this revision. PatchSiren helps prioritize defensive review and does not prove exposure or remediation on any system.