CVE-2025-9906
CVE-2025-9906 is a high-severity vulnerability in Keras with a CVSS 3.x base score of 7.3. It is not currently listed as actively exploited by CISA, and its EPSS exploit-prediction score is low. The underlying weakness is classified as CWE-502.
Key facts
- Severity: High (CVSS 3.x base score 7.3)
- CVSS v4: 8.6
- EPSS exploit prediction: 0% (10th percentile)
- Actively exploited: Not listed in CISA KEV
- EU (EUVD) id: EUVD-2025-30281
- Weakness: CWE-502
- Affected product: Keras
- Published:
- Last modified:
Description
The Keras Model.load_model method can be exploited to achieve arbitrary code execution, even with safe_mode=True. One can create a specially crafted .keras model archive that, when loaded via Model.load_model, will trigger arbitrary code to be executed. This is achieved by crafting a special config.json (a file within the .keras archive) that will invoke keras.config.enable_unsafe_deserialization() to disable safe mode. Once safe mode is disable, one can use the Lambda layer feature of keras, which allows arbitrary Python code in the form of pickled code. Both can appear in the same archive. Simply the keras.config.enable_unsafe_deserialization() needs to appear first in the archive and the Lambda with arbitrary code needs to be second.
Frequently asked questions
- What is CVE-2025-9906?
- The Keras Model.load_model method can be exploited to achieve arbitrary code execution, even with safe_mode=True. One can create a specially crafted .keras model archive that, when loaded via Model.load_model, will trigger arbitrary code to be executed. This is achieved by crafting a special config.json (a file within the .keras archive) that will invoke keras.config.enable_unsafe_deserialization() to disable safe mode. Once safe mode is disable, one can use the Lambda layer feature of keras, which allows arbitrary Python code in the form of pickled code. Both can appear in the same archive. Simply the keras.config.enable_unsafe_deserialization() needs to appear first in the archive and the Lambda with arbitrary code needs to be second.
- How severe is CVE-2025-9906?
- CVE-2025-9906 has a CVSS 3.x base score of 7.3, rated high severity. It is exploitable over local access with low attack complexity, requires low privileges and user interaction. Impact on confidentiality is high, integrity high, and availability high.
- Is CVE-2025-9906 being actively exploited?
- It is not currently listed in CISA's KEV catalog. Its EPSS exploit-prediction score is 0% (10th percentile), an estimate of the probability of exploitation in the next 30 days.
- What products are affected by CVE-2025-9906?
- CVE-2025-9906 affects Keras. See the affected-products list for the exact vulnerable versions.
- How do I fix CVE-2025-9906?
- Review the linked vendor and NVD advisories for patched versions and mitigations, then upgrade or apply the recommended workaround. Given its high severity, prioritise patching exposed systems.
- Does CVE-2025-9906 have an EU (EUVD) identifier?
- Yes. CVE-2025-9906 is tracked in the ENISA EU Vulnerability Database (EUVD) as EUVD-2025-30281.
- When was CVE-2025-9906 published?
- CVE-2025-9906 was published on 2025-09-19 and last updated on 2026-06-17.
References
Affected products (1)
- cpe:2.3:a:keras:keras:*:*:*:*:*:*:*:*
More vulnerabilities in Keras
- CVE-2026-12481 — Critical (CVSS 9.8): A vulnerability in keras-team/keras version 3.14.0 allows for arbitrary code execution due to improper handling of…
- CVE-2025-1550 — Critical (CVSS 9.8): The Keras Model.load_model function permits arbitrary code execution, even with safe_mode=True, through a manually…
- CVE-2024-3660 — Critical (CVSS 9.8): A arbitrary code injection vulnerability in TensorFlow's Keras framework (<2.13) allows attackers to execute arbitrary…
- CVE-2026-11816 — High (CVSS 8.1): Keras versions prior to 3.14.0 are vulnerable to a path traversal issue in the archive extraction utilities located in…
- CVE-2026-1462 — High (CVSS 7.8): A vulnerability in the `TFSMLayer` class of the `keras` package, version 3.13.0, allows attacker-controlled TensorFlow…
- CVE-2025-8747 — High (CVSS 7.8): A safe mode bypass vulnerability in the `Model.load_model` method in Keras versions 3.0.0 through 3.10.0 allows an…
Other CWE-502 (Deserialization of Untrusted Data) vulnerabilities
- CVE-2026-69836 — Critical (CVSS 10.0): Deserialization of untrusted data in Microsoft Entra ID allows an unauthorized attacker to execute code over a network.
- CVE-2026-17061 — Critical (CVSS 10.0): A Deserialization of Untrusted Data vulnerability affecting SIMULIA Execution Engine from Release 2023 through Release…
- CVE-2026-11756 — Critical (CVSS 10.0): A Deserialization of Untrusted Data vulnerability affecting Station Launcher App in 3DEXPERIENCE platform from Release…
- CVE-2026-41104 — Critical (CVSS 10.0): Deserialization of untrusted data in Microsoft Planetary Computer Pro allows an unauthorized attacker to disclose…
- CVE-2026-43633 — Critical (CVSS 10.0): HestiaCP versions 1.9.0 through 1.9.4 contain a deserialization vulnerability in the web terminal component caused by a…
- CVE-2026-33819 — Critical (CVSS 10.0): Deserialization of untrusted data in Microsoft Bing allows an unauthorized attacker to execute code over a network.
Browse all CWE-502 (Deserialization of Untrusted Data) vulnerabilities →