Threats Tagged 'cve-2026-12570'

Keras versions up to 3.15.0 contain a vulnerability in the keras.models.load_model() function that allows denial of service (DoS) via loading specially crafted .keras model files. The flaw arises from lack of validation on dataset shape or size in H5IOStore.__getitem__, leading to unbounded memory allocation and potential out-of-memory termination. This vulnerability bypasses a previous fix for a related issue (CVE-2026-0897). The attack vector includes poisoned models from untrusted sources, posing a risk to machine learning pipelines that load external models.

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.