CVE-2025-55556
Google Tensorflow 2.18.0
Raw vector
CVSS:3.1/AV:N/AC:L/PR:N/UI:N/S:U/C:N/I:L/A:LSummary
CVE-2025-55556 is a medium-severity Embedded Malicious Code (CWE-506) vulnerability in Google Tensorflow. Its CVSS base score is 6.5 (Medium).
Operationally, exploitation aligns with the MITRE ATT&CK technique Embedded Payloads (T1027.009); ranked at the 6th percentile by exploit likelihood (below the median); it is not currently listed in the CISA KEV catalog; a public proof-of-concept is referenced.
This vulnerability is AI-related — categorised as Deep Learning Frameworks; in the Other ATLAS/OWASP Terms risk domain.
The strongest mitigations our analysis identified map to SA-11 (Developer Testing and Evaluation) and SC-44 (Detonation Chambers) — see the control section below for these in your framework.
OWASP Top 10 for Web (2025)
EU & UK References
- 🇪🇺 ENISA EUVD: EUVD-2025-31129
Vulnerability Data
TensorFlow v2.18.0 was discovered to output random results when compiling Embedding, leading to unexpected behavior in the application.
- CWE(s)
AI Security AnalysisAI
- AI Category
- Deep Learning Frameworks
- Risk Domain
- Other ATLAS/OWASP Terms
- OWASP Top 10 for LLMs 2025
- None mapped
- Classification Reason
- Matched keywords: tensorflow
Related Threats
MITRE ATT&CK Enterprise Techniques
CVEs Like This One
Affected Assets
Mitigating Controls
Control response
—
—
- 3 hardening rules · 2 OS baselines
—
Mitigating Controls (NIST 800-53 r5) AI
Developer testing and evaluation can discover embedded malicious code before release.
Detonation chambers can reveal embedded malicious code through controlled execution analysis.
Integrity verification tools directly detect unauthorized or malicious code insertions.
Tamper-resistance and detection mechanisms identify malicious code introduced via the supply chain.
Component authenticity requirements reduce the chance malicious code is embedded by developers or suppliers.
Component authenticity controls stop introduction of malicious or counterfeit code into the product.
Mitigating Controls (NIST CSF 2.0) AI
Derived directly from the weakness types (CWEs) cited in the NVD entry via our AI-authored CWE→CSF cross-walk (authority under review) — links open the control.
Assessing authenticity and integrity before acquisition catches embedded malicious code.
Secure SDLC practices directly prevent introduction of malicious code during development.
Due diligence prior to supplier relationships helps avoid sources of embedded malicious code.
Supplier risk assessment reduces likelihood of receiving products containing malicious code.
Pre-acquisition assessment of critical suppliers mitigates risk of malicious code in procured software.
Mitigating Controls (ISO/IEC 27001:2022 Annex A) AI
Derived directly from the weakness types (CWEs) cited in the NVD entry via our AI-authored CWE→ISO cross-walk (authority under review) — links open the control.
Requiring suppliers to furnish component lists, attestations, and cryptographic verification of delivered artefacts makes it harder for an attacker to embed hidden malicious code that would otherwise go undetected through the supply chain.
Checks that executed code has not been tampered with and monitoring for malware-associated activity reduce the likelihood that hidden malicious code remains active.
Banning unapproved or unknown software and requiring testing plus authorization reduces the chance that hidden malicious code will be introduced into production environments.
Hardening callouts derived
Configuration rules from DISA STIG baselines that bear on weaknesses of the type cited by this CVE. Each rule is shown with the relationship its mapping actually records, against the CWE it was authored against. Derived via CVE→CWE over `controls_xwalks` (authoritative rows only; rows rated `none` are excluded).
Oracle Linux 9 (1 rule)
- V-271452 OL 9 must use a Linux Security Module configured to enforce limits on system services. prevents CWE-506
RHEL 9 (1 rule)
- V-258078 RHEL 9 must use a Linux Security Module configured to enforce limits on system services. prevents CWE-506