Security Bulletin: Arbitrary Code Execution in Keras

Summary Keras is used by many machine learning frameworks and applications as part of their deep learning infrastructure. Remote attackers can execute arbitrary code, leading to full system compromise, data breaches, and potential lateral movement where the identified vulnerability is present...

📄 Keras 2.15 Insecure Deserialization

Keras version 2.15 insecure deserialization proof of concept exploit. A security issue in certain versions of Keras allows attackers to craft a malicious model file typically a .keras or HDF5-based model containing unsafe serialization primitives. When such a model is loaded, the deserialization...

CVE-2025-33212

NVIDIA NeMo Framework contains a vulnerability in model loading that could allow an attacker to exploit improper control mechanisms if a user loads a maliciously crafted file. A successful exploit of this vulnerability might lead to code execution, escalation of privileges, denial of service, and...

EUVD-2025-203816

NVIDIA NeMo Framework contains a vulnerability in model loading that could allow an attacker to exploit improper control mechanisms if a user loads a maliciously crafted file. A successful exploit of this vulnerability might lead to code execution, escalation of privileges, denial of service, and...

CVE-2025-33212

NVIDIA NeMo Framework contains a vulnerability in model loading that could allow an attacker to exploit improper control mechanisms if a user loads a maliciously crafted file. A successful exploit of this vulnerability might lead to code execution, escalation of privileges, denial of service, and...

CVE-2025-33212

NVIDIA NeMo Framework contains a vulnerability in model loading that could allow an attacker to exploit improper control mechanisms if a user loads a maliciously crafted file. A successful exploit of this vulnerability might lead to code execution, escalation of privileges, denial of service, and...

CVE-2025-33212

NVIDIA NeMo Framework contains a vulnerability in model loading that could allow an attacker to exploit improper control mechanisms if a user loads a maliciously crafted file. A successful exploit of this vulnerability might lead to code execution, escalation of privileges, denial of service, and...

CVE-2025-33212

NVIDIA NeMo Framework contains a vulnerability in model loading that could allow an attacker to exploit improper control mechanisms if a user loads a maliciously crafted file. A successful exploit of this vulnerability might lead to code execution, escalation of privileges, denial of service, and...

NVIDIA Nemo Framework 代码问题漏洞

NVIDIA Nemo Framework is a framework for building and deploying generative AI models from NVIDIA. The NVIDIA NeMo Framework contains a security vulnerability that can be exploited by attackers to cause code execution, elevation of privilege, denial of service, and data tampering...

PT-2025-51759

NVIDIA NeMo Framework contains a vulnerability in model loading that could allow an attacker to exploit improper control mechanisms if a user loads a maliciously crafted file. A successful exploit of this vulnerability might lead to code execution, escalation of privileges, denial of service, and...

Server-Side Request Forgery (SSRF)

Keras is vulnerable to Server-Side Request Forgery SSRF. The vulnerability is due to improper handling of the StringLookup layer during model loading from a crafted .keras archive, which allows an attacker to supply local or remote file paths as vocabulary inputs and exploit tf.io.gfile behavior ...

CVE-2025-66448

A remote code execution vulnerability has been identified in vLLM. An attacker can exploit a weakness in the model loading process to silently fetch and run unauthorized, malicious Python code on the host system. This happens because the engine mistakenly executes code from a remote repository...

OESA-2025-2690 python-Keras security update

Keras is a high-level neural networks API for Python. Security Fixes: The Keras.Model.loadmodel method, including when executed with the intended security mitigation safemode=True, is vulnerable to arbitrary local file loading and Server-Side Request Forgery SSRF. This vulnerability stems from th...

OESA-2025-2691 python-Keras security update

Keras is a high-level neural networks API for Python. Security Fixes: The Keras.Model.loadmodel method, including when executed with the intended security mitigation safemode=True, is vulnerable to arbitrary local file loading and Server-Side Request Forgery SSRF. This vulnerability stems from th...

OESA-2025-2689 python-Keras security update

Keras is a high-level neural networks API for Python. Security Fixes: The Keras.Model.loadmodel method, including when executed with the intended security mitigation safemode=True, is vulnerable to arbitrary local file loading and Server-Side Request Forgery SSRF. This vulnerability stems from th...

CVE-2025-12487

oobabooga text-generation-webui trustremotecode Reliance on Untrusted Inputs Remote Code Execution Vulnerability. This vulnerability allows remote attackers to execute arbitrary code on affected installations of oobabooga text-generation-webui. Authentication is not required to exploit this...

CVE-2025-12488

oobabooga text-generation-webui trustremotecode Reliance on Untrusted Inputs Remote Code Execution Vulnerability. This vulnerability allows remote attackers to execute arbitrary code on affected installations of oobabooga text-generation-webui. Authentication is not required to exploit this...

CVE-2025-12487 oobabooga text-generation-webui trust_remote_code Reliance on Untrusted Inputs Remote Code Execution Vulnerability

oobabooga text-generation-webui trustremotecode Reliance on Untrusted Inputs Remote Code Execution Vulnerability. This vulnerability allows remote attackers to execute arbitrary code on affected installations of oobabooga text-generation-webui. Authentication is not required to exploit this...

Arbitrary Code Execution

Keras is vulnerable to Arbitrary Code Execution. The vulnerability is due to Model.loadmodel not honoring safemode=True when reading legacy .h5/.hdf5 archives and deserializing pickled Lambda-layer code from a crafted model file, which allows an attacker to supply a malicious archive that execute...

EUVD-2025-36634

Keras is vulnerable to arbitrary local file loading and Server-Side Request Forgery...