ByteDance DeerFlow versions prior to commit 92c7a20 contain a sandbox escape vulnerability in bash tool handling that allows attackers to execute arbitrary commands on the host system by bypassing regex-based validation using shell features such as directory changes and relative paths. Attackers can exploit the incomplete shell semantics modeling to read and modify files outside the sandbox boundary and achieve arbitrary command execution through subprocess invocation with shell interpretation enabled.

Hugging Face Transformers MaskFormer Model Deserialization of Untrusted Data Remote Code Execution Vulnerability. This vulnerability allows remote attackers to execute arbitrary code on affected installations of Hugging Face Transformers. User interaction is required to exploit this vulnerability in that the target must visit a malicious page or open a malicious file. The specific flaw exists within the parsing of model files. The issue results from the lack of proper validation of user-supplied data, which can result in deserialization of untrusted data. An attacker can leverage this vulnerability to execute code in the context of the current user. Was ZDI-CAN-25191.

Deserialization of Untrusted Data vulnerability in BoldGrid weForms weforms allows Object Injection.This issue affects weForms: from n/a through <= 1.6.26.

SvelteKit is a web development framework. The SvelteKit framework offers developers an option to create simple REST APIs. This is done by defining a `+server.js` file, containing endpoint handlers for different HTTP methods. SvelteKit provides out-of-the-box cross-site request forgery (CSRF) protection to its users. While the implementation does a sufficient job in mitigating common CSRF attacks, prior to version 1.15.1, the protection can be bypassed by simply specifying a different `Content-Type` header value. If abused, this issue will allow malicious requests to be submitted from third-party domains, which can allow execution of operations within the context of the victim's session, and in extreme scenarios can lead to unauthorized access to users’ accounts. SvelteKit 1.15.1 updates the `is_form_content_type` function call in the CSRF protection logic to include `text/plain`. As additional hardening of the CSRF protection mechanism against potential method overrides, SvelteKit 1.15.1 is now performing validation on `PUT`, `PATCH` and `DELETE` methods as well. This latter hardening is only needed to protect users who have put in some sort of `?_method= override` feature themselves in their `handle` hook, so that the request that resolve sees could be `PUT`/`PATCH`/`DELETE` when the browser issues a `POST` request.

An issue was discovered in Spipu HTML2PDF before 5.2.4. Attackers can trigger deserialization of arbitrary data via the injection of a malicious <link> tag in the converted HTML document.

Cross-Site Request Forgery (CSRF), Deserialization of Untrusted Data vulnerability in ThemeHigh Job Manager & Career – Manage job board listings, and recruitments.This issue affects Job Manager & Career – Manage job board listings, and recruitments: from n/a through 1.4.4.

PyTorch-Lightning versions 2.6.0 and earlier contain an insecure deserialization vulnerability (CWE-502) in the checkpoint loading mechanism. The LightningModule.load_from_checkpoint() method, which is commonly used to load saved model states, internally calls torch.load() without setting the security-restrictive weights_only=True parameter. This default behavior allows the deserialization of arbitrary Python objects via the Pickle module. A remote attacker can exploit this by providing a maliciously crafted checkpoint file, leading to arbitrary code execution on the victim's system when the file is loaded.

The snorkel library thru v0.10.0 contains an insecure deserialization vulnerability (CWE-502) in the Trainer.load() method of the Trainer class. The method loads model checkpoint files using torch.load() without enabling the security-restrictive weights_only=True parameter. This default behavior allows the deserialization of arbitrary Python objects via the Pickle module. A remote attacker can exploit this by providing a maliciously crafted model file, leading to arbitrary code execution on the victim's system when the file is loaded via the vulnerable method.

The snorkel library thru v0.10.0 contains a critical insecure deserialization vulnerability (CWE-502) in the BaseLabeler.load() method of the BaseLabeler class. The method loads serialized labeler models using the unsafe pickle.load() function on user-supplied file paths without any validation or security controls. Python's pickle module is inherently dangerous for deserializing untrusted data, as it can execute arbitrary code during the deserialization process. A remote attacker can exploit this by providing a maliciously crafted pickle file, leading to arbitrary code execution on the victim's system when the file is loaded via the vulnerable method.

The snorkel library thru v0.10.0 contains an insecure deserialization vulnerability (CWE-502) in the MultitaskClassifier.load() method of the MultitaskClassifier class. The method loads model weight files using torch.load() without enabling the security-restrictive weights_only=True parameter. This default behavior allows the deserialization of arbitrary Python objects via the Pickle module. A remote attacker can exploit this by providing a maliciously crafted model file, leading to arbitrary code execution on the victim's system when the file is loaded via the vulnerable method.

The CosyVoice project thru commit 6e01309e01bc93bbeb83bdd996b1182a81aaf11e (2025-30-21) contains an insecure deserialization vulnerability (CWE-502) in its model loading process. When loading model files (.pt) from a user-specified directory (via the --model_dir argument), the code uses torch.load() without the security-restrictive weights_only=True parameter. This allows the deserialization of arbitrary Python objects via the Pickle module. An attacker can exploit this by providing a maliciously crafted model directory containing .pt files with embedded pickle payloads. When a victim loads this directory using CosyVoice's web interface, the malicious payload is executed, leading to remote code execution on the victim's system.

The Clone plugin for WordPress is vulnerable to PHP Object Injection in all versions up to, and including, 2.4.6 via deserialization of untrusted input in the 'recursive_unserialized_replace' function. This makes it possible for unauthenticated attackers to inject a PHP Object. No known POP chain is present in the vulnerable software. If a POP chain is present via an additional plugin or theme installed on the target system, it could allow the attacker to delete arbitrary files, retrieve sensitive data, or execute code.

PyTorch is a Python package that provides tensor computation. Prior to version 2.10.0, a vulnerability in PyTorch's `weights_only` unpickler allows an attacker to craft a malicious checkpoint file (`.pth`) that, when loaded with `torch.load(..., weights_only=True)`, can corrupt memory and potentially lead to arbitrary code execution. Version 2.10.0 fixes the issue.

The Better Search Replace plugin for WordPress is vulnerable to PHP Object Injection in all versions up to, and including, 1.4.4 via deserialization of untrusted input. This makes it possible for unauthenticated attackers to inject a PHP Object. No POP chain is present in the vulnerable plugin. If a POP chain is present via an additional plugin or theme installed on the target system, it could allow the attacker to delete arbitrary files, retrieve sensitive data, or execute code.

NVIDIA BioNeMo contains a vulnerability where a user could cause a deserialization of untrusted data. A successful exploit of this vulnerability might lead to code execution, denial of service, information disclosure, and data tampering.

Lexmark Markvision Enterprise before 2.3.0 misuses the Apache Commons Collections Library, leading to remote code execution because of Java deserialization.

IBM Security Identity Manager 6.0.0 could allow a remote attacker to execute arbitrary code on the system, caused by the deserialization of untrusted data. By persuading a victim to visit a specially crafted Web site, an attacker could exploit this vulnerability to execute arbitrary code on the system. IBM X-Force ID: 166456.

Redisson is a Java Redis client that uses the Netty framework. Prior to version 3.22.0, some of the messages received from the Redis server contain Java objects that the client deserializes without further validation. Attackers that manage to trick clients into communicating with a malicious server can include especially crafted objects in its responses that, once deserialized by the client, force it to execute arbitrary code. This can be abused to take control of the machine the client is running in. Version 3.22.0 contains a patch for this issue. Some post-fix advice is available. Do NOT use `Kryo5Codec` as deserialization codec, as it is still vulnerable to arbitrary object deserialization due to the `setRegistrationRequired(false)` call. On the contrary, `KryoCodec` is safe to use. The fix applied to `SerializationCodec` only consists of adding an optional allowlist of class names, even though making this behavior the default is recommended. When instantiating `SerializationCodec` please use the `SerializationCodec(ClassLoader classLoader, Set<String> allowedClasses)` constructor to restrict the allowed classes for deserialization.

Deserialization of Untrusted Data vulnerability in Kleor Contact Manager contact-manager allows Object Injection.This issue affects Contact Manager: from n/a through <= 9.1.1.

LMDeploy is a toolkit for compressing, deploying, and serving LLMs. Prior to version 0.11.1, an insecure deserialization vulnerability exists in lmdeploy where torch.load() is called without the weights_only=True parameter when loading model checkpoint files. This allows an attacker to execute arbitrary code on the victim's machine when they load a malicious .bin or .pt model file. This issue has been patched in version 0.11.1.

cryptidy through 1.2.4 allows code execution via untrusted data because pickle.loads is used. This occurs in aes_decrypt_message in symmetric_encryption.py.

The value function in jsonpath 1.1.1 lib/index.js is vulnerable to Prototype Pollution.

A vulnerability has been identified in the UA.Testclient utility, which is included in Rexroth IndraWorks. All versions prior to 15V24 are affected. This flaw allows an attacker to execute arbitrary code on the user's system by parsing a manipulated file containing malicious serialized data. Exploitation requires user interaction, specifically opening a specially crafted file, which then causes the application to deserialize the malicious data, enabling Remote Code Execution (RCE). This can lead to a complete compromise of the system running the UA.Testclient.

The affected Cognex product, the In-Sight OPC Server versions v5.7.4 (96) and prior, deserializes untrusted data, which could allow a remote attacker access to system level permission commands and local privilege escalation.

A vulnerability has been identified in Rexroth IndraWorks. This flaw allows an attacker to execute arbitrary code on the user's system by parsing a manipulated file containing malicious serialized data. Exploitation requires user interaction, specifically opening a specially crafted file, which then causes the application to deserialize the malicious data, enabling Remote Code Execution (RCE). This can lead to a complete compromise of the system running Rexroth IndraWorks.

A vulnerability has been identified in Rexroth IndraWorks. This flaw allows an attacker to execute arbitrary code on the user's system by parsing a manipulated file containing malicious serialized data. Exploitation requires user interaction, specifically opening a specially crafted file, which then causes the application to deserialize the malicious data, enabling Remote Code Execution (RCE). This can lead to a complete compromise of the system running Rexroth IndraWorks.

A vulnerability has been identified in the OPC.Testclient utility, which is included in Rexroth IndraWorks. All versions prior to 15V24 are affected. This flaw allows an attacker to execute arbitrary code on the user's system by parsing a manipulated file containing malicious serialized data. Exploitation requires user interaction, specifically opening a specially crafted file, which then causes the application to deserialize the malicious data, enabling Remote Code Execution (RCE). This can lead to a complete compromise of the system running the OPC.Testclient.

Deserialization of Untrusted Data vulnerability in sbouey Falang multilanguage falang allows Object Injection.This issue affects Falang multilanguage: from n/a through <= 1.3.65.

MONAI (Medical Open Network for AI) is an AI toolkit for health care imaging. In versions up to and including 1.5.0, the `pickle_operations` function in `monai/data/utils.py` automatically handles dictionary key-value pairs ending with a specific suffix and deserializes them using `pickle.loads()` . This function also lacks any security measures. The deserialization may lead to code execution. As of time of publication, no known fixed versions are available.

Datart 1.0.0-rc.3 is vulnerable to Directory Traversal. The configuration file handling of the application allows attackers to upload arbitrary YAML files to the config/jdbc-driver-ext.yml path. The application parses this file using SnakeYAML's unsafe load() or loadAs() method without input sanitization. This allows deserialization of attacker-controlled YAML content, leading to arbitrary class instantiation. Under certain conditions, this can be exploited to achieve remote code execution (RCE).

The Starter Templates by Kadence WP WordPress plugin before 1.2.17 unserialises the content of an imported file, which could lead to PHP object injection issues when an admin import (intentionally or not) a malicious file and a suitable gadget chain is present on the blog.

The ImageMagick Engine plugin for WordPress is vulnerable to deserialization of untrusted input via the 'cli_path' parameter in versions up to, and including 1.7.5. This makes it possible for unauthenticated users to call files using a PHAR wrapper, granted they can trick a site administrator into performing an action such as clicking on a link, that will deserialize and call arbitrary PHP Objects that can be used to perform a variety of malicious actions granted a POP chain is also present. It also requires that the attacker is successful in uploading a file with the serialized payload.

The WPtouch WordPress plugin before 4.3.45 unserialises the content of an imported settings file, which could lead to PHP object injections issues when an user import (intentionally or not) a malicious settings file and a suitable gadget chain is present on the blog.

OperaMasks SDK ELite Script Engine v0.5.0 was discovered to contain a deserialization vulnerability.

NVIDIA Merlin Transformers4Rec for Linux contains a vulnerability in the Trainer component, where a user could cause a deserialization issue. A successful exploit of this vulnerability might lead to code execution, denial of service, information disclosure, and data tampering.

NVIDIA NVTabular for Linux contains a vulnerability in the Workflow component, where a user could cause a deserialization issue. A successful exploit of this vulnerability might lead to code execution, denial of service, information disclosure, and data tampering.

Jooby is a web framework for Java and Kotlin. The pac4j io.jooby.internal.pac4j.SessionStoreImpl#get module deserializes untrusted data. This vulnerability is fixed in 2.17.0 (2.x) and 3.7.0 (3.x).

A remote code execution vulnerability exists in "Microsoft COM for Windows" when it fails to properly handle serialized objects, aka "Microsoft COM for Windows Remote Code Execution Vulnerability." This affects Windows 7, Windows Server 2012 R2, Windows RT 8.1, Windows Server 2008, Windows Server 2012, Windows 8.1, Windows Server 2016, Windows Server 2008 R2, Windows 10, Windows 10 Servers.

In the LibreOffice 7-1 series in versions prior to 7.1.2, and in the 7-0 series in versions prior to 7.0.5, the denylist can be circumvented by manipulating the link so it doesn't match the denylist but results in ShellExecute attempting to launch an executable type.

By launching the drb_remote_codeexec exploit, a Metasploit Framework user will inadvertently expose Metasploit to the same deserialization issue that is exploited by that module, due to the reliance on the vulnerable Distributed Ruby class functions. Since Metasploit Framework typically runs with elevated privileges, this can lead to a system compromise on the Metasploit workstation. Note that an attacker would have to lie in wait and entice the Metasploit user to run the affected module against a malicious endpoint in a "hack-back" type of attack. Metasploit is only vulnerable when the drb_remote_codeexec module is running. In most cases, this cannot happen automatically.

IBM InfoSphere Information Server 11.3, 11.5, and 11.7 could allow a remote attacker to execute arbitrary code on the system, caused by the deserialization of untrusted data. By persuading a victim to visit a specially crafted Web site, an attacker could exploit this vulnerability to execute arbitrary code on the system. IBM X-Force ID: 176677.

FasterXML jackson-databind 2.x before 2.9.10.8 mishandles the interaction between serialization gadgets and typing, related to org.apache.tomcat.dbcp.dbcp2.datasources.PerUserPoolDataSource.

FasterXML jackson-databind 2.x before 2.9.10.8 mishandles the interaction between serialization gadgets and typing, related to org.apache.commons.dbcp2.cpdsadapter.DriverAdapterCPDS.

FasterXML jackson-databind 2.x before 2.9.10.8 mishandles the interaction between serialization gadgets and typing, related to org.apache.tomcat.dbcp.dbcp.cpdsadapter.DriverAdapterCPDS.

FasterXML jackson-databind 2.x before 2.9.10.8 mishandles the interaction between serialization gadgets and typing, related to oadd.org.apache.commons.dbcp.cpdsadapter.DriverAdapterCPDS.

FasterXML jackson-databind 2.x before 2.9.10.8 mishandles the interaction between serialization gadgets and typing, related to org.apache.tomcat.dbcp.dbcp2.cpdsadapter.DriverAdapterCPDS.

FasterXML jackson-databind 2.x before 2.9.10.4 mishandles the interaction between serialization gadgets and typing, related to javax.swing.JEditorPane.

FasterXML jackson-databind 2.x before 2.9.10.4 mishandles the interaction between serialization gadgets and typing, related to org.apache.commons.proxy.provider.remoting.RmiProvider (aka apache/commons-proxy).

FasterXML jackson-databind 2.x before 2.9.10.4 mishandles the interaction between serialization gadgets and typing, related to org.aoju.bus.proxy.provider.remoting.RmiProvider (aka bus-proxy).