Finding and attribution
When rpm -q -p or a similar manifest-processing query/verify flow handles an attacker-supplied manifest, macro expansion of manifest entries can execute shell commands with the privileges of the rpm process. User or automation must process the crafted input; Red Hat documents --nomanifest as a mitigation.
Product: RPM. AI system: AISLE AI-native analyzer (specific model/version not disclosed). Team / organization: AISLE in partnership with Red Hat.
CVE identifiers: CVE-2026-95519.
AI and human contribution (source account): Red Hat’s CVE page explicitly says the issue was discovered by AISLE in partnership with Red Hat. AISLE’s itemized CVE page says “Discovered by AISLE.” Specific model/version and human/model work split are undisclosed.
Human involvement is only described where the source account establishes it; unknown involvement does not mean none. Vendor confirmation of a flaw does not confirm every claim about AI contribution.
Discovery context: The Red Hat CNA publication and AISLE case page are dated September 24, 2026. Red Hat confirms the flaw and AI-finder credit. The public Red Hat bug record says no released package fix was established at its report stage and upstream was not notified; current public fix status was not established in this check. Red Hat documents a mitigation. This is not an in-the-wild claim.
Remediation status: not established. Red Hat recommends avoiding rpm queries on untrusted files, using --nomanifest, or validating manifest entries. Public evidence checked here does not establish a released upstream/package fix; do not interpret the mitigation as a fix.
A source-documented fix does not establish installed patches or current safety. Source checking does not independently reproduce the finding or establish observed exploitation.
Disclosed 24 Sep 2026. AI attribution dated 24 Sep 2026. Sources checked 15 Nov 2026; these are separate dates, not a live installation check.
Recorded CVSS: High · highest recorded score 7.8. Documented technical effects: code execution, data disclosure, data modification, denial of service.
Severity, consequences and exploitation evidence
Case grouping uses the highest recorded CVSS score, with its version-specific severity band. It is descriptive, not a combined or editorial risk index. All recorded assessments remain below; scores from different versions and sources may disagree. CVSS specifications define version-specific scoring and severity bands.
- CVE-2026-95519: CVSS 3.1 · 7.8 (High) ·
CVSS:3.1/AV:L/AC:L/PR:N/UI:R/S:U/C:H/I:H/A:H. redhat (CNA; CVE Program record) · checked 16 Nov 2026.
CVE-2026-95519: code execution, data disclosure, data modification, denial of service.
Prerequisites / affected scope: A user or automation must process a crafted RPM manifest via `rpm -q -p` or a similar flow. Embedded shell commands execute with the rpm process privileges; the CNA says successful exploitation can fully compromise confidentiality, integrity, and availability of the affected account..
Exploitation evidence: Exploitation in the wild is not established by the cited description.. Impact source · checked 16 Nov 2026.
Documented fix: Red Hat recommends avoiding rpm queries on untrusted files, using --nomanifest, or validating manifest entries. Public evidence checked here does not establish a released upstream/package fix; do not interpret the mitigation as a fix.