CWE-502: Deserialization of Untrusted Data vulnerabilities

NVIDIA Transformers4Rec contains a deserialization vulnerability (CWE-502) that allows improper handling of untrusted data. Exploitation could lead to denial of service but does not impact confidentiality or integrity. The vulnerability has a medium severity score of 4.3 and currently no known exploits or patches are reported.

A denial of service vulnerability could be triggered by sending specially crafted HTTP requests to server function endpoints, this could lead to excessive CPU usage; affecting the following packages: react-server-dom-webpack, react-server-dom-parcel, react-server-dom-turbopack (versions 19.0.0 through 19.0.7, 19.1.0 through 19.1.8, and 19.2.0 through 19.2.7).

Deserialization of untrusted data vulnerability that may allow class-registration checks to be bypassed during Java lambda deserialization. Only lambda capture class is affected This issue affects Apache Fory: from before 1.4.0. Users are recommended to upgrade to version 1.4.0, which fixes the issue.

Heap type confusion and out-of-bounds read/write in the Apache Fory C++ implementation. When deserializing data in compatible mode, the field-skip paths do not correctly validate the declared field types against the actual data, so input with an inconsistent schema can cause type confusion and out-of-bounds memory access. Only the C++ implementation is affected; other language implementations of Apache Fory are not. This issue affects Apache Fory C++: from 0.14.0 before 1.4.0. Users are recommended to upgrade to version 1.4.0, which fixes the issue.

CVE-2026-12484 is a high-severity vulnerability in keras-team/keras version 3.15.0 that allows unsafe deserialization of attacker-controlled PyTorch pickle data via the public method keras.layers.TorchModuleWrapper.from_config. This method calls torch.load without requiring an explicit unsafe opt-in or enforcing safe deserialization, potentially leading to arbitrary code execution if untrusted Keras layer configurations are processed. The vulnerability exists because safe deserialization is only enforced when explicitly guarded by Keras safe mode, which is not the default behavior.

IBM Langflow OSS 1.0.0 through 1.10.0 contain a critical remote code execution vulnerability in the disk-based caching mechanism. The AsyncDiskCache class uses Python's unsafe pickle.loads() function to deserialize cached objects from disk without validation, integrity verification, or authentication, enabling arbitrary code execution when malicious pickle payloads are processed. Attackers who can influence cached data through file system access, malicious workflow inputs, custom components, or API manipulation can achieve complete system compromise with the privileges of the Langflow server process.

The Uncanny Automator plugin for WordPress up to version 7.3.1.4 contains a deserialization vulnerability that allows unauthenticated attackers to delete arbitrary files on the server. This occurs due to insufficient file path validation in the fr_token function. Exploitation requires a Forminator form linked to an Uncanny Automator recipe configured to allow submissions from 'Everyone', enabling malicious serialized payloads to trigger the vulnerability. The plugin includes a gadget chain internally, so no external libraries are needed for exploitation. This can lead to remote code execution if critical files like wp-config.php are deleted.

NVIDIA TensorRT contains a deserialization vulnerability (CWE-502) that could allow an attacker to cause deserialization of untrusted data. Exploiting this vulnerability might lead to code execution. The vulnerability has a medium severity with a CVSS score of 5.3. No specific affected versions or patches are currently identified.

NVIDIA TensorRT-LLM contains a deserialization vulnerability in its visual gen server component. This vulnerability arises from unsafe deserialization of unauthorized zeroMQ data, potentially allowing an attacker to execute arbitrary code. The issue affects version 0.0 of TensorRT-LLM. The CVSS score is 6.4, indicating a medium severity level.

NVIDIA TensorRT-LLM has a deserialization vulnerability in its inter-process communication layer that can be exploited by a local attacker with the same user privileges. Successful exploitation could result in code execution, information disclosure, data tampering, and denial of service. The vulnerability is identified as CWE-502 and carries a high severity rating with a CVSS score of 7.8. No patch or official remediation guidance is currently available.

NVIDIA TensorRT-LLM for Linux has a deserialization vulnerability in its restricted unpickler used for model weight deserialization. This flaw allows a local, unauthenticated attacker to deserialize untrusted data, potentially leading to code execution, privilege escalation, data tampering, and information disclosure. The vulnerability is identified as CVE-2026-24233 with a high severity rating and a CVSS score of 8.4. No official patch or remediation guidance has been provided yet by the vendor.

Deserialization of untrusted data in .NET allows an unauthorized attacker to execute code locally.

Protection mechanism failure in .NET Framework allows an unauthorized attacker to execute code locally.

Deserialization of untrusted data in Microsoft Dynamics NAV allows an unauthorized attacker to execute code over a network.

Deserialization of untrusted data in Windows Wireless Wide Area Network Service allows an authorized attacker to elevate privileges locally.

Deserialization of untrusted data in Azure Active Directory allows an unauthorized attacker to deny service over a network.

Deserialization of untrusted data in Microsoft Office SharePoint allows an unauthorized attacker to execute code over a network.

Deserialization of untrusted data in Microsoft Office SharePoint allows an unauthorized attacker to execute code over a network.

Deserialization of untrusted data in SQL Server allows an unauthorized attacker to execute code over a network.