Skip to content
Fencer Intel
critical 9.8

CVE-2019-6446 in the critical feed json

Numpy Deserialization of Untrusted Data

published 7y ago modified 4mo ago also GHSA-9fq2-x9r6-wfmf , PYSEC-2019-108 , EUVD-2019-0098

cvss
9.8
version 3.0, NVD
cisa kev
Not listed
not in CISA's catalog
epss
27.3%
EPSS 27.3% chance of exploitation in 30 days, percentile 98%
public exploits
0
none known

Summary

An issue was discovered in NumPy before 1.16.3. It uses the pickle Python module unsafely, which allows remote attackers to execute arbitrary code via a crafted serialized object, as demonstrated by a numpy.load call. NOTE: third parties dispute this issue because it is a behavior that might have legitimate applications in (for example) loading serialized Python object arrays from trusted and authenticated sources.

Patch / fix

  • pypi numpy 1.16.1, 1.16.3
  • pytz RHEL 8 2017.2-12.module+el8.1.0+3111+de3f2d8e
  • babel RHEL 8 2.5.1-9.module+el8.1.0+3111+de3f2d8e
  • numpy RHEL 8 1:1.14.2-13.module+el8.1.0+3323+7ac3e00f
  • scipy RHEL 8 1.0.0-20.module+el8.1.0+3323+7ac3e00f
  • Cython RHEL 8 0.28.1-7.module+el8.1.0+3111+de3f2d8e
  • pytest RHEL 8 3.4.2-13.module+el8.1.0+3111+de3f2d8e
  • PyYAML RHEL 8 3.12-16.module+el8.1.0+3111+de3f2d8e
  • python2 RHEL 8 2.7.16-12.module+el8.1.0+4148+33a50073

red hat advisories RHSA-2019:3335RHSA-2019:3704

  • python-numpy SLE Module for Python 2 15 SP1, SLE Module for Basesystem 15 SP1, SLE Module for Basesystem 15 1.16.1-4.8.1
  • python-numpy_1_16_1-gnu-hpc SLE Module for HPC 15 SP1, SLE Module for HPC 15 1.16.1-4.8.1
  • python-numpy SLE Point of Sale 11 SP3 1.3.0-1.3.3.1
  • python-numpy_1_13_3-gnu-hpc SLE Module for HPC 12 1.13.3-4.9.1
  • python-numpy SLES for SAP 12 SP4, SLES for SAP 12 SP3, SLES 12 SP4 +5 more 1.8.0-5.8.1
  • python-numpy SLE Module for Basesystem 15 1.14.0-4.5.1
  • python-numpy_1_14_0-gnu-hpc SLE Module for HPC 15 1.14.0-4.5.1
  • python-numpy SLES for SAP 11 SP4, SLES 11 SP4, SLE Software Development Kit 11 SP4 1.8.0-6.4.1

suse advisories SUSE-SU-2019:2462-2SUSE-SU-2019:2462-1SUSE-SU-2019:13977-1SUSE-SU-2019:0448-1SUSE-SU-2019:0419-1SUSE-SU-2019:0418-1

  • python-numpy openSUSE Tumbleweed 2.4.6-1.1
  • python-numpy1 openSUSE Tumbleweed 1.26.4-1.1
  • python-numpy openSUSE Tumbleweed 1.21.2-1.1
  • python-numpy openSUSE Tumbleweed 1.26.2-5.1
  • python-numpy Package Hub 15 SP1 1.16.1-bp151.2.3.1
  • python-numpy_1_16_1-gnu-hpc Package Hub 15 SP1 1.16.1-bp151.2.3.1
  • python-numpy openSUSE Leap 15.1 1.16.1-lp151.5.3.1
  • python-numpy_1_16_1-gnu-hpc openSUSE Leap 15.1 1.16.1-lp151.5.3.1

opensuse advisories openSUSE-SU-2026:11371-1openSUSE-SU-2024:14311-1openSUSE-SU-2024:11243-1openSUSE-SU-2024:13820-1openSUSE-SU-2019:2259-1openSUSE-SU-2019:2227-1

  • python2-py AlmaLinux 8 1.5.3-6.module_el8.6.0+2781+fed64c13
  • python2-dns AlmaLinux 8 1.15.0-10.module_el8.6.0+2781+fed64c13
  • python2-idna AlmaLinux 8 2.5-7.module_el8.6.0+2781+fed64c13
  • python2-mock AlmaLinux 8 2.0.0-13.module_el8.6.0+2781+fed64c13
  • python2-pytz AlmaLinux 8 2017.2-12.module_el8.6.0+2781+fed64c13
  • python2-attrs AlmaLinux 8 17.4.0-10.module_el8.6.0+2781+fed64c13
  • python2-Cython AlmaLinux 8 0.28.1-7.module_el8.6.0+2781+fed64c13
  • python2-pluggy AlmaLinux 8 0.6.0-8.module_el8.6.0+2781+fed64c13

almalinux advisories ALSA-2019:3335

  • numpy Rocky Linux 8 1:1.14.3-9.el8
  • pytz Rocky Linux 8 2017.2-12.module+el8.5.0+706+735ec4b3
  • pytz Rocky Linux 8 2017.2-12.module+el8.4.0+403+9ae17a31
  • babel Rocky Linux 8 2.5.1-9.module+el8.4.0+403+9ae17a31
  • scipy Rocky Linux 8 1.0.0-20.module+el8.3.0+120+426d8baf
  • Cython Rocky Linux 8 0.28.1-7.module+el8.5.0+706+735ec4b3
  • Cython Rocky Linux 8 0.28.1-7.module+el8.4.0+403+9ae17a31
  • pytest Rocky Linux 8 3.4.2-13.module+el8.5.0+706+735ec4b3

rocky linux advisories RLSA-2019:3704RLSA-2019:3335

Is CVE-2019-6446 in your dependencies? Fencer shows which of your repositories are affected as soon as an advisory lands, and opens the fix as a pull request.

Check with Fencer

Vendor advisories

What the vendors themselves say about their products: affected, fixed or not affected, and what to do.

red hat product security vex CVE-2019-6446 updated cvss 8.8

numpy: crafted serialized object passed in numpy.load() in pickle python module allows arbitrary code execution

  • Red Hat Ceph Storage 4 numpy
    affected: no fix available
  • Red Hat Enterprise Linux 6 numpy, numpy-f2py Red Hat Enterprise Linux 7 numpy, numpy-f2py Red Hat OpenStack Platform 13 (Queens) python2-numpy, python2-numpy-f2py Red Hat OpenStack Platform 14 (Rocky) python2-numpy, python2-numpy-f2py Red Hat Software Collections python27-numpy, rh-python35-numpy, rh-python36-numpy, python27-numpy-f2py, rh-python35-numpy-f2py, rh-python36-numpy-f2py Red Hat Virtualization 4 rhvm-appliance
    affected: no fix planned “Will not fix”
  • Red Hat Enterprise Linux AppStream (v. 8) numpy 1.14.3-9.el8, python3-numpy 1.14.3-9.el8, python3-numpy-f2py 1.14.3-9.el8, babel 2.5.1-9.module+el8.1.0+3111+de3f2d8e, pytz 2017.2-12.module+el8.1.0+3111+de3f2d8e, PyYAML 3.12-16.module+el8.1.0+3111+de3f2d8e, +87 more
    fixed: vendor fix RHSA-2019:3335

Red Hat CSAF VEX © Red Hat, Inc., from https://security.access.redhat.com/data/csaf/v2/vex/; changed: reduced to product statuses and remediations. CC BY 4.0 · csaf

Exploit prediction

27.27% chance of exploitation in the next 30 days, higher than 98% of all scored CVEs (FIRST EPSS, 2026-11-09).

Exploits and checks

  • bugzilla.suse.com https://bugzilla.suse.com/show_bug.cgi?id=1122208 referenced PoC, tagged "exploit" by a source, unverified
  • github.com https://github.com/numpy/numpy/issues/12759 referenced PoC, tagged "exploit" by a source, unverified

CVSS scores

assigned by version score vector
NVD 3.0 9.8 CVSS:3.0/AV:N/AC:L/PR:N/UI:N/S:U/C:H/I:H/A:H
NVD 2.0 7.5 AV:N/AC:L/Au:N/C:P/I:P/A:P
GitHub 3.0 9.8 CVSS:3.0/AV:N/AC:L/PR:N/UI:N/S:U/C:H/I:H/A:H

Affected packages

ecosystem package affected fixed in
pypi numpy
<1.16.3
<1.16.1
1.16.3
1.16.1

Classification

Weakness
  • CWE-502 Deserialization of Untrusted Data

References

Sources

Fencer Intel keeps one entry per vulnerability: the CVE id wins, then GHSA, and records that alias each other are merged, except that two CVEs always keep separate entries. The same data is available as JSON.