Cyber Resilience

CVE-2025-26644

Microsoft Windows 10 1809 ≤ 10.0.17763.7136

Published
08 April 2025
Modified
10 July 2025
Patch / advisory
CVSS Score v3.1 5.1
Click a component to see what it means
Raw vectorCVSS:3.1/AV:L/AC:H/PR:N/UI:N/S:U/C:N/I:H/A:N
EPSS Score 0.0053 42th percentile
Risk Priority 41 floored blend · peak EPSS

Summary

CVE-2025-26644 is a medium-severity Inadequate Detection or Handling of Adversarial Input Perturbations in Automated Recognition Mechanism (CWE-1039) vulnerability in Microsoft Windows 10 1809. Its CVSS base score is 5.1 (Medium).

Operationally, ranked at the 42th percentile by exploit likelihood (below the median); it is not currently listed in the CISA KEV catalog.

The strongest mitigations our analysis identified map to SA-11 (Developer Testing and Evaluation) and SI-10 (Information Input Validation) — see the control section below for these in your framework.

EU & UK References

Vulnerability Data

Automated recognition mechanism with inadequate detection or handling of adversarial input perturbations in Windows Hello allows an unauthorized attacker to perform spoofing locally.

CWE(s)

Related Threats

MITRE ATLAS TechniquesAI

MITRE ATLAS techniques

AML.T0008.003: Physical CountermeasuresAML.T0015: Evade AI ModelAML.T0017.000: Adversarial AI AttacksAML.T0031: Erode AI Model IntegrityAML.T0043: Craft Adversarial DataAML.T0043.000: White-Box Optimization

CVEs Like This One

CVE-2024-49073Same product: Microsoft Windows 10 1809
CVE-2024-49081Same product: Microsoft Windows 10 1809
CVE-2024-49098Same product: Microsoft Windows 10 1809
CVE-2024-49078Same product: Microsoft Windows 10 1809
CVE-2024-49083Same product: Microsoft Windows 10 1809
CVE-2024-49087Same product: Microsoft Windows 10 1809
CVE-2024-49101Same product: Microsoft Windows 10 1809
CVE-2024-49110Same product: Microsoft Windows 10 1809
CVE-2024-49099Same product: Microsoft Windows 10 1809
CVE-2024-49094Same product: Microsoft Windows 10 1809

Affected Assets

microsoft
windows 10 1809
≤ 10.0.17763.7136 · ≤ 10.0.17763.7136
microsoft
windows 10 21h2
≤ 10.0.19044.5737
microsoft
windows 10 22h2
≤ 10.0.19045.5737
microsoft
windows 11 22h2
≤ 10.0.22621.5189
microsoft
windows 11 23h2
≤ 10.0.22631.5189
microsoft
windows 11 24h2
≤ 10.0.26100.3775
microsoft
windows server 2019
≤ 10.0.17763.7136
microsoft
windows server 2025
≤ 10.0.26100.3775

Mitigating Controls

Mitigating Controls (NIST 800-53 r5) AI

Developer testing can specifically exercise the recognition mechanism against adversarial examples to surface the weakness.

Input validation can reject or sanitize perturbed inputs before they reach the recognition mechanism.

System monitoring can observe anomalous classifications or input patterns indicative of adversarial perturbations.

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.

DE.CM-09 mostly match
prevents

Monitoring runtime data and inputs directly supports detection of adversarial perturbations to recognition mechanisms.

PR.PS-06 mostly match
prevents

Secure SDLC practices include robustness testing and adversarial-input handling during model development.

ID.RA-01 partial match
prevents

Recording this class of ML vulnerability is a prerequisite for subsequent mitigation.

ID.RA-03 partial match
prevents

Explicit identification of adversarial ML threats informs the need to address this weakness.

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.

finds

Security testing in development can explicitly cover adversarial input testing for recognition mechanisms.

mitigates

Threat intelligence can surface adversarial ML attack patterns but does not implement technical controls.

prevents

Secure development lifecycle requires adversarial testing and robustness validation for ML-based recognition components.

prevents

Application security requirements can mandate detection and handling of adversarial perturbations in automated recognition systems.

prevents

Secure architecture principles include resilience against input manipulation and adversarial ML attacks.

prevents

Secure coding practices can incorporate input sanitization and adversarial robustness checks for ML models.

References