Vulnerabilities (CVE)

Join the Common Vulnerabilities and Exposures (CVE) community and start to get notified about new vulnerabilities.

Filtered by CWE-502
Total 934 CVE
CVE Vendors Products Updated CVSS v2 CVSS v3
CVE-2021-24040 1 Facebook 1 Parlai 2021-09-23 7.5 HIGH 9.8 CRITICAL
Due to use of unsafe YAML deserialization logic, an attacker with the ability to modify local YAML configuration files could provide malicious input, resulting in remote code execution or similar risks. This issue affects ParlAI prior to v1.1.0.
CVE-2021-39207 1 Facebook 1 Parlai 2021-09-23 6.5 MEDIUM 8.8 HIGH
parlai is a framework for training and evaluating AI models on a variety of openly available dialogue datasets. In affected versions the package is vulnerable to YAML deserialization attack caused by unsafe loading which leads to Arbitary code execution. This security bug is patched by avoiding unsafe loader users should update to version above v1.1.0. If upgrading is not possible then users can change the Loader used to SafeLoader as a workaround. See commit 507d066ef432ea27d3e201da08009872a2f37725 for details.
CVE-2021-36766 1 Concretecms 1 Concrete Cms 2021-09-22 6.5 MEDIUM 7.2 HIGH
Concrete5 through 8.5.5 deserializes Untrusted Data. The vulnerable code is located within the controllers/single_page/dashboard/system/environment/logging.php Logging::update_logging() method. User input passed through the logFile request parameter is not properly sanitized before being used in a call to the file_exists() PHP function. This can be exploited by malicious users to inject arbitrary PHP objects into the application scope (PHP Object Injection via phar:// stream wrapper), allowing them to carry out a variety of attacks, such as executing arbitrary PHP code.
CVE-2021-34394 1 Nvidia 9 Jetson Agx Xavier 16gb, Jetson Agx Xavier 32gb, Jetson Agx Xavier 8gb and 6 more 2021-09-20 4.6 MEDIUM 6.7 MEDIUM
Trusty contains a vulnerability in the NVIDIA OTE protocol that is present in all TAs. An incorrect message stream deserialization allows an attacker to use the malicious CA that is run by the user to cause the buffer overflow, which may lead to information disclosure and data modification.
CVE-2021-29200 1 Apache 1 Ofbiz 2021-09-20 7.5 HIGH 9.8 CRITICAL
Apache OFBiz has unsafe deserialization prior to 17.12.07 version An unauthenticated user can perform an RCE attack
CVE-2021-30128 1 Apache 1 Ofbiz 2021-09-20 10.0 HIGH 9.8 CRITICAL
Apache OFBiz has unsafe deserialization prior to 17.12.07 version
CVE-2021-37579 1 Apache 1 Dubbo 2021-09-17 7.5 HIGH 9.8 CRITICAL
The Dubbo Provider will check the incoming request and the corresponding serialization type of this request meet the configuration set by the server. But there's an exception that the attacker can use to skip the security check (when enabled) and reaching a deserialization operation with native java serialization. Apache Dubbo 2.7.13, 3.0.2 fixed this issue by quickly fail when any unrecognized request was found.
CVE-2021-26295 1 Apache 1 Ofbiz 2021-09-16 7.5 HIGH 9.8 CRITICAL
Apache OFBiz has unsafe deserialization prior to 17.12.06. An unauthenticated attacker can use this vulnerability to successfully take over Apache OFBiz.
CVE-2021-36163 1 Apache 1 Dubbo 2021-09-14 7.5 HIGH 9.8 CRITICAL
In Apache Dubbo, users may choose to use the Hessian protocol. The Hessian protocol is implemented on top of HTTP and passes the body of a POST request directly to a HessianSkeleton: New HessianSkeleton are created without any configuration of the serialization factory and therefore without applying the dubbo properties for applying allowed or blocked type lists. In addition, the generic service is always exposed and therefore attackers do not need to figure out a valid service/method name pair. This is fixed in 2.7.13, 2.6.10.1
CVE-2019-18631 1 Centrify 3 Authentication Service, Infrastructure Services, Privilege Elevation Service 2021-09-13 5.1 MEDIUM 7.8 HIGH
The Windows component of Centrify Authentication and Privilege Elevation Services 3.4.0, 3.4.1, 3.4.2, 3.4.3, 3.5.0, 3.5.1 (18.8), 3.5.2 (18.11), and 3.6.0 (19.6) does not properly handle an unspecified exception during use of partially trusted assemblies to serialize input data, which allows attackers to execute arbitrary code inside the Centrify process via (1) a crafted application that makes a pipe connection to the process and sends malicious serialized data or (2) a crafted Microsoft Management Console snap-in control file.
CVE-2021-32568 1 Mrdoc 1 Mrdoc 2021-09-09 6.8 MEDIUM 7.8 HIGH
mrdoc is vulnerable to Deserialization of Untrusted Data
CVE-2021-36231 1 Unit4 1 Mik.starlight 2021-09-08 9.0 HIGH 8.8 HIGH
Deserialization of untrusted data in multiple functions in MIK.starlight 7.9.5.24363 allows authenticated remote attackers to execute operating system commands by crafting serialized objects.
CVE-2018-6162 4 Apple, Debian, Google and 1 more 6 Macos, Debian Linux, Chrome and 3 more 2021-09-08 6.8 MEDIUM 8.8 HIGH
Improper deserialization in WebGL in Google Chrome on Mac prior to 68.0.3440.75 allowed a remote attacker to potentially exploit heap corruption via a crafted HTML page.
CVE-2021-21677 1 Jenkins 1 Code Coverage Api 2021-09-08 6.5 MEDIUM 8.8 HIGH
Jenkins Code Coverage API Plugin 1.4.0 and earlier does not apply Jenkins JEP-200 deserialization protection to Java objects it deserializes from disk, resulting in a remote code execution vulnerability.
CVE-2021-39132 1 Pagerduty 1 Rundeck 2021-09-08 6.5 MEDIUM 8.8 HIGH
Rundeck is an open source automation service with a web console, command line tools and a WebAPI. Prior to version 3.3.14 and version 3.4.3, an authorized user can upload a zip-format plugin with a crafted plugin.yaml, or a crafted aclpolicy yaml file, or upload an untrusted project archive with a crafted aclpolicy yaml file, that can cause the server to run untrusted code on Rundeck Community or Enterprise Edition. An authenticated user can make a POST request, that can cause the server to run untrusted code on Rundeck Enterprise Edition. The zip-format plugin issues requires authentication and authorization to these access levels, and affects all Rundeck editions:`admin` level access to the `system` resource type. The ACL Policy yaml file upload issues requires authentication and authorization to these access levels, and affects all Rundeck editions: `create` `update` or `admin` level access to a `project_acl` resource, and/or`create` `update` or `admin` level access to the `system_acl` resource. The unauthorized POST request requires authentication, but no specific authorization, and affects Rundeck Enterprise only. Patches are available in versions 3.4.3, 3.3.14
CVE-2021-34066 1 Edgegallery 1 Developer-be 2021-09-07 10.0 HIGH 9.8 CRITICAL
An issue was discovered in EdgeGallery/developer before v1.0. There is a "Deserialization of yaml file" vulnerability that can allow attackers to execute system command through uploading the malicious constructed YAML file.
CVE-2021-21741 1 Zte 2 Zxv10 M910, Zxv10 M910 Firmware 2021-09-07 7.5 HIGH 9.8 CRITICAL
A conference management system of ZTE is impacted by a command execution vulnerability. Since the soapmonitor's java object service is enabled by default, the attacker could exploit this vulnerability to execute arbitrary commands by sending a deserialized payload to port 5001.
CVE-2021-24579 1 Bold-themes 1 Bold Page Builder 2021-09-02 6.5 MEDIUM 8.8 HIGH
The bt_bb_get_grid AJAX action of the Bold Page Builder WordPress plugin before 3.1.6 passes user input into the unserialize() function without any validation or sanitisation, which could lead to a PHP Object Injection. Even though the plugin did not contain a suitable gadget to fully exploit the issue, other installed plugins on the blog could allow such issue to be exploited and lead to RCE in some cases.
CVE-2021-38585 1 Cpanel 1 Cpanel 2021-08-20 6.5 MEDIUM 7.2 HIGH
The WHM Locale Upload feature in cPanel before 98.0.1 allows unserialization attacks (SEC-585).
CVE-2021-23420 1 Codeception 1 Codeception 2021-08-19 10.0 HIGH 9.8 CRITICAL
This affects the package codeception/codeception from 4.0.0 and before 4.1.22, before 3.1.3. The RunProcess class can be leveraged as a gadget to run arbitrary commands on a system that is deserializing user input without validation.