CVE-2025-65886
Published: 28 January 2026
Summary
CVE-2025-65886 is a high-severity Uncontrolled Resource Consumption (CWE-400) vulnerability in Oneflow Oneflow. Its CVSS base score is 7.5 (High).
Operationally, exploitation aligns with the MITRE ATT&CK technique Application or System Exploitation (T1499.004); ranked at the 40.0th percentile by exploit likelihood (below the median); it is not currently listed in the CISA KEV catalog; a public proof-of-concept is referenced.
The strongest mitigations our analysis identified are NIST 800-53 SC-5 (Denial-of-service Protection) and SI-10 (Information Input Validation).
Deeper analysis
CVE-2025-65886 is a shape mismatch vulnerability affecting OneFlow version 0.9.0, an open-source framework handling tensor operations. The flaw enables attackers to trigger a Denial of Service (DoS) condition by supplying crafted tensor shapes, as documented under CWE-400 (Uncontrolled Resource Consumption). Published on 2026-01-28 with a CVSS v3.1 base score of 7.5 (AV:N/AC:L/PR:N/UI:N/S:U/C:N/I:N/A:H), it poses a high-impact availability risk without compromising confidentiality or integrity.
The vulnerability is exploitable by unauthenticated attackers over the network with low attack complexity and no user interaction required. By delivering specially crafted tensor shapes to a vulnerable OneFlow instance, attackers can cause resource exhaustion or crashes, leading to service disruption. This scenario targets deployments exposing OneFlow endpoints, such as in distributed training or inference setups.
Mitigation details and patches are referenced in the OneFlow GitHub repository (https://github.com/Oneflow-Inc/oneflow), specific issue tracker (https://github.com/Oneflow-Inc/oneflow/issues/10666), and related sites including http://oneflow.com and https://github.com/Daisy2ang. Security practitioners should review these for upgrade guidance or workarounds to address the issue in v0.9.0.
As a tensor shape issue in an AI/ML framework, CVE-2025-65886 highlights risks in deep learning pipelines where malformed inputs can propagate to DoS, though no real-world exploitation has been reported in the provided data.
EU & UK References
- 🇪🇺 ENISA EUVD: EUVD-2025-206484
Vulnerability details
A shape mismatch vulnerability in OneFlow v0.9.0 allows attackers to cause a Denial of Service (DoS) via supplying crafted tensor shapes.
- CWE(s)
Related Threats
MITRE ATT&CK Enterprise TechniquesAI
Why these techniques?
Directly enables Endpoint DoS via application exploitation (crafted tensor input triggers resource exhaustion in exposed OneFlow service).
CVEs Like This One
Affected Assets
Mitigating Controls
Mitigating Controls (NIST 800-53 r5) AI
Validates incoming tensor shapes to block crafted inputs that trigger shape mismatch and resource exhaustion in OneFlow.
Implements DoS protections such as rate limiting or resource quotas to mitigate network-based resource consumption attacks via malformed tensor shapes.
Requires timely patching of the OneFlow v0.9.0 shape mismatch flaw as referenced in the GitHub issue tracker to eliminate the vulnerability.