CVE-2026-12570 is a medium-severity (CVSS 5.5) CWE-770 vulnerability in keras. A fix is available for keras — see the affected versions and patch details below.
Denial of Service via HDF5 Shape Bomb in keras.models.load_model() in keras-team/keras
Exploitation Status
Proof-of-concept exploit code exists
- CISA’s SSVC triage found public proof-of-concept exploit code for this CVE, though no confirmed active exploitation.
Exploitation and automatability from CISA’s SSVC triage for CVE-2026-12570.
EPSS Exploitation Probability
Probability of exploitation in the next 30 days, from FIRST.org EPSS.
How urgent is this, really
CVE-2026-12570 by exploitation likelihood (EPSS) against impact (CVSS). Outside the shaded patch-first corner.
Where this sits among everything scored
Of 393,867 CVEs with a current EPSS score, this one falls in the < 10% band (highlighted). Counts from FIRST.org, log-scaled.
Real-World Exposure
kerasReal-time download stats are indexed for npm and PyPI packages. This vulnerability affects PyPI packages — download data is not available via public APIs for these ecosystems.
Description
A vulnerability in keras-team/keras versions <= 3.15.0 allows for a denial of service (DoS) attack when loading malicious .keras model files via the keras.models.load_model() function. The H5IOStore.getitem method in keras/src/saving/saving_lib.py does not validate the shape or size of datasets, leading to unbounded memory allocation. A specially crafted .keras file can exploit this flaw to trigger an out-of-memory (OOM) condition, causing the process to be terminated (exit code 137). This issue bypasses the fix for CVE-2026-0897, which only addressed a similar vulnerability in KerasFileEditor. The attack vector includes poisoned models from public repositories or malicious model registries, posing a risk to machine learning pipelines that process untrusted models.
Affected Packages
| Ecosystem | Package | Vulnerable range | Fix |
|---|---|---|---|
| 🐍PyPI | keras | all versions | 3.15.0pip install --upgrade 'keras==3.15.0' |
Detection & mitigation playbook
Open-source dependencyDetect
Scan your dependency tree (package-lock.json, pnpm-lock.yaml, requirements.txt, go.sum, etc.) for keras, including transitive dependencies — a direct dependency you never call can still pull in a vulnerable version.
Fix
Update keras to 3.15.0 or later, then make sure no transitive (indirect) dependency still pins the vulnerable range — O3 confirms CVE-2026-12570 is resolved across your whole dependency graph.
Workarounds
Cap what an attacker can consume: apply request size, rate and timeout limits in front of the affected component, and run it with memory and CPU limits so exhaustion degrades one worker rather than the whole service.
Fixing This On Your OS
If you run this on a Linux distribution, patch through your package manager against the distro's own security advisory below — it tracks the exact backported fix for your release, which can ship on a different timeline (and sometimes a different severity) than the upstream project.
Red Hat ships Keras across multiple OpenShift AI (RHOAI) model-serving and workbench images: KServe agent, controller, router, and storage-initializer images, the ModelMesh runtime adapter image, TensorFlow pipeline-runtime images, and Jupyter TensorFlow workbench images (CUDA and ROCm variants). The…
Only load .keras model files from trusted sources; do not load user-supplied or externally-sourced model files without validation. Apply pod-level memory limits/requests to constrain the blast radius of an out-of-memory condition until an upstream fix is available.Source: Red Hat security advisory for CVE-2026-12570 (CC BY 4.0)
Frequently Asked Questions
Is CVE-2026-12570 in your dependencies?
Find it across PyPI, including transitive dependencies.