CVE-2026-66007
Datasets Path Traversal via Unsanitized file_name Metadata
Datasets through 5.0.0, fixed in commit f989ef9, contains a path traversal vulnerability in folder-based dataset builders where the file_name metadata field is not properly validated before being joined to the dataset directory. Attackers can supply crafted file_name values with directory traversal sequences to read arbitrary local files, which are then embedded into output when save_to_disk or push_to_hub is called.
Attack profile
How easy it is to exploit
- Attack vector
- Network
- Complexity
- Low
- Privileges needed
- None
- User interaction
- Active
What an attacker can affect
- Confidentiality
- High
- Integrity
- None
- Availability
- None
Exploit likelihood
What each source reports
CVE List Score used
6.9
Datasets Path Traversal via Unsanitized file_name Metadata
Datasets through 5.0.0, fixed in commit f989ef9, contains a path traversal vulnerability in folder-based dataset builders where the file_name metadata field is not properly validated before being joined to the dataset directory. Attackers can supply crafted file_name values with directory traversal sequences to read arbitrary local files, which are then embedded into output when save_to_disk or push_to_hub is called.
- Record
CVE-2026-66007- CVSS 4.0
- 6.9 Medium
CVSS:4.0/AV:N/AC:L/AT:N/PR:N/UI:A/VC:H/VI:N/VA:N/SC:N/SI:N/SA:N - CVSS 3.1
- 6.5 Medium
CVSS:3.1/AV:N/AC:L/PR:N/UI:R/S:U/C:H/I:N/A:N - Weakness
- CWE-22
- Updated
- 2026-11-02
EUVDEUVD-2026-48613
6.9
Datasets through 5.0.0, fixed in commit f989ef9, contains a path traversal vulnerability in folder-based dataset builders where the file_name metadata field is not properly validated before being joined to the dataset directory. Attackers can supply crafted file_name values with directory traversal sequences to read arbitrary local files, which are then embedded into output when save_to_disk or push_to_hub is called.
- Record
EUVD-2026-48613- CVSS 4.0
- 6.9 Medium
CVSS:4.0/AV:N/AC:L/AT:N/PR:N/UI:A/VC:H/VI:N/VA:N/SC:N/SI:N/SA:N - Updated
- 2026-11-02
GitHub AdvisoriesGHSA-379c-qx7v-6h59
6.9
Hugging Face Datasets folder-based builders allow path traversal through file_name metadata
Datasets through 5.0.0, fixed in f989ef9, contains a path traversal vulnerability in folder-based dataset builders where the file_name metadata field is not properly validated before being joined to the dataset directory. Attackers can supply crafted file_name values with directory traversal sequences to read arbitrary local files, which are then embedded into output when save_to_disk or push_to_hub is called.
- Record
GHSA-379c-qx7v-6h59- CVSS 4.0
- 6.9 Medium
CVSS:4.0/AV:N/AC:L/AT:N/PR:N/UI:A/VC:H/VI:N/VA:N/SC:N/SI:N/SA:N - CVSS 3.1
- 6.5 Medium
CVSS:3.1/AV:N/AC:L/PR:N/UI:R/S:U/C:H/I:N/A:N - Weakness
- CWE-22
- Updated
- 2026-11-02
NVD
6.9
Datasets through 5.0.0, fixed in commit f989ef9, contains a path traversal vulnerability in folder-based dataset builders where the file_name metadata field is not properly validated before being joined to the dataset directory. Attackers can supply crafted file_name values with directory traversal sequences to read arbitrary local files, which are then embedded into output when save_to_disk or push_to_hub is called.
- Record
CVE-2026-66007- CVSS 4.0
- 6.9 Medium
CVSS:4.0/AV:N/AC:L/AT:N/PR:N/UI:A/VC:H/VI:N/VA:N/SC:N/SI:N/SA:N/E:X/CR:X/IR:X/AR:X/MAV:X/MAC:X/MAT:X/MPR:X/MUI:X/MVC:X/MVI:X/MVA:X/MSC:X/MSI:X/MSA:X/S:X/AU:X/R:X/V:X/RE:X/U:X - CVSS 3.1
- 6.5 Medium
CVSS:3.1/AV:N/AC:L/PR:N/UI:R/S:U/C:H/I:N/A:N - Weakness
- CWE-22
- Updated
- 2026-11-02
OSV
6.9
Datasets Path Traversal via Unsanitized file_name Metadata
Datasets through 5.0.0, fixed in commit f989ef9, contains a path traversal vulnerability in folder-based dataset builders where the file_name metadata field is not properly validated before being joined to the dataset directory. Attackers can supply crafted file_name values with directory traversal sequences to read arbitrary local files, which are then embedded into output when save_to_disk or push_to_hub is called.
- Record
CVE-2026-66007- CVSS 4.0
- 6.9 Medium
CVSS:4.0/AV:N/AC:L/AT:N/PR:N/UI:A/VC:H/VI:N/VA:N/SC:N/SI:N/SA:N - Weakness
- CWE-22
- Updated
- 2026-11-02
OSVGHSA-379c-qx7v-6h59
6.9
Hugging Face Datasets folder-based builders allow path traversal through file_name metadata
Datasets through 5.0.0, fixed in f989ef9, contains a path traversal vulnerability in folder-based dataset builders where the file_name metadata field is not properly validated before being joined to the dataset directory. Attackers can supply crafted file_name values with directory traversal sequences to read arbitrary local files, which are then embedded into output when save_to_disk or push_to_hub is called.
- Record
GHSA-379c-qx7v-6h59- CVSS 4.0
- 6.9 Medium
CVSS:4.0/AV:N/AC:L/AT:N/PR:N/UI:A/VC:H/VI:N/VA:N/SC:N/SI:N/SA:N - CVSS 3.1
- 6.5 Medium
CVSS:3.1/AV:N/AC:L/PR:N/UI:R/S:U/C:H/I:N/A:N - Weakness
- CWE-22
- Updated
- 2026-11-02
OSVPYSEC-2026-3716
6.9
Datasets through 5.0.0, fixed in commit f989ef9, contains a path traversal vulnerability in folder-based dataset builders where the file_name metadata field is not properly validated before being joined to the dataset directory. Attackers can supply crafted file_name values with directory traversal sequences to read arbitrary local files, which are then embedded into output when save_to_disk or push_to_hub is called.
- Record
PYSEC-2026-3716- CVSS 4.0
- 6.9 Medium
CVSS:4.0/AV:N/AC:L/AT:N/PR:N/UI:A/VC:H/VI:N/VA:N/SC:N/SI:N/SA:N/E:X/CR:X/IR:X/AR:X/MAV:X/MAC:X/MAT:X/MPR:X/MUI:X/MVC:X/MVI:X/MVA:X/MSC:X/MSI:X/MSA:X/S:X/AU:X/R:X/V:X/RE:X/U:X - Updated
- 2026-11-02
Affected products
Affected packages
| Package | Affected | Fixed |
|---|---|---|
datasetsPyPI |
before 5.0.1 | 5.0.1 |
datasetsPyPI |
before 5.0.1, plus listed versions 0.0.91.0.01.0.11.0.21.1.0 Show 45 moreShow less1.1.11.1.21.1.31.10.01.10.11.10.21.11.01.12.01.12.11.13.01.13.11.13.21.13.31.14.01.15.01.15.11.16.01.16.11.17.01.18.01.18.11.18.21.18.31.18.41.2.01.2.11.3.01.4.01.4.11.5.01.6.01.6.11.6.21.7.01.8.01.9.02.0.02.1.02.10.02.10.12.11.02.12.02.13.02.13.12.13.2 |
5.0.1 |
History
| Recorded | Changed | Sources |
|---|---|---|
| first indexed | CVE List, EUVD, GitHub Advisories, NVD, OSV |
Dates are when Every CVE recorded each change. Upstream dates may differ.
In short
CVE-2026-66007 is a medium vulnerability (CVSS 6.9) in Huggingface Datasets, published July 24, 2026. No source lists public exploit code. EPSS puts the chance of exploitation in the next 30 days at 0.55%, the 46th percentile. Fixed in datasets 5.0.1.
References
github.com/huggingface/datasets/issues/8324
github.com/huggingface/datasets/pull/8325
github.com/huggingface/datasets/commit/f989ef9b4cc6c0039a7a82458eebca49e2b58b4b
vulncheck.com/advisories/datasets-path-traversal-via-unsanitized-file-name-metadata
nvd.nist.gov/vuln/detail/CVE-2026-66007
github.com/pypa/advisory-database/tree/main/vulns/datasets/PYSEC-2026-3716.yaml
github.com/advisories/GHSA-379c-qx7v-6h59
github.com/CVEProject/cvelistV5/tree/main/cves/2026/66xxx/CVE-2026-66007.json
github.com/huggingface/datasets
Left out here: 68 versions. The JSON record has them.