Apache Camel RabbitMQ enables Java deserialization by default. Apache Camel 2.22.x, 2.23.x, 2.24.x, 2.25.0, 3.0.0 up to 3.1.0 are affected. 2.x users should upgrade to 2.25.1, 3.x users should upgrade to 3.2.0.
An issue was found in Apache Airflow versions 1.10.10 and below. When using CeleryExecutor, if an attack can connect to the broker (Redis, RabbitMQ) directly, it was possible to insert a malicious payload directly to the broker which could lead to a deserialization attack (and thus remote code execution) on the Worker.
A deserialization vulnerability existed in dubbo 2.7.5 and its earlier versions, which could lead to malicious code execution. Most Dubbo users use Hessian2 as the default serialization/deserialization protool, during Hessian2 deserializing the HashMap object, some functions in the classes stored in HasMap will be executed after a series of program calls, however, those special functions may cause remote command execution. For example, the hashCode() function of the EqualsBean class in rome-1.7.0.jar will cause the remotely load malicious classes and execute malicious code by constructing a malicious request. This issue was fixed in Apache Dubbo 2.6.9 and 2.7.8.
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.
Apache Camel Netty enables Java deserialization by default. Apache Camel 2.22.x, 2.23.x, 2.24.x, 2.25.0, 3.0.0 up to 3.1.0 are affected. 2.x users should upgrade to 2.25.1, 3.x users should upgrade to 3.2.0.
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.
An untrusted deserialization was found in the org.apache.xmlrpc.parser.XmlRpcResponseParser:addResult method of Apache XML-RPC (aka ws-xmlrpc) library. A malicious XML-RPC server could target a XML-RPC client causing it to execute arbitrary code. Apache XML-RPC is no longer maintained and this issue will not be fixed.
Apache Tika before 1.14 allows Java code execution for serialized objects embedded in MATLAB files. The issue exists because Tika invokes JMatIO to do native deserialization.
Included in Log4j 1.2 is a SocketServer class that is vulnerable to deserialization of untrusted data which can be exploited to remotely execute arbitrary code when combined with a deserialization gadget when listening to untrusted network traffic for log data. This affects Log4j versions up to 1.2 up to 1.2.17.
Apache Olingo versions 4.0.0 to 4.6.0 provide the AbstractService class, which is public API, uses ObjectInputStream and doesn't check classes being deserialized. If an attacker can feed malicious metadata to the class, then it may result in running attacker's code in the worse case.
Deserialization of untrusted data in python in pyfory versions 0.12.0 through 0.12.2, or the legacy pyfury versions from 0.1.0 through 0.10.3: allows arbitrary code execution. An application is vulnerable if it reads pyfory serialized data from untrusted sources. An attacker can craft a data stream that selects pickle-fallback serializer during deserialization, leading to the execution of `pickle.loads`, which is vulnerable to remote code execution. Users are recommended to upgrade to pyfory version 0.12.3 or later, which has removed pickle fallback serializer and thus fixes this issue.
The java.io.ObjectInputStream is known to cause Java serialisation issues. This issue here is exposed by the "webtools/control/httpService" URL, and uses Java deserialization to perform code execution. In the HttpEngine, the value of the request parameter "serviceContext" is passed to the "deserialize" method of "XmlSerializer". Apache Ofbiz is affected via two different dependencies: "commons-beanutils" and an out-dated version of "commons-fileupload" Mitigation: Upgrade to 16.11.06 or manually apply the commits from OFBIZ-10770 and OFBIZ-10837 on branch 16
Manipulating classpath asset file URLs, an attacker could guess the path to a known file in the classpath and have it downloaded. If the attacker found the file with the value of the tapestry.hmac-passphrase configuration symbol, most probably the webapp's AppModule class, the value of this symbol could be used to craft a Java deserialization attack, thus running malicious injected Java code. The vector would be the t:formdata parameter from the Form component.
Arbitrary Class Instantiation via Model Manifest in Apache OpenNLP ExtensionLoader Versions Affected: before 1.9.5, before 2.5.9, before 3.0.0-M3 Description: The ExtensionLoader.instantiateExtension(Class, String) method loads a class by its fully-qualified name via Class.forName() and invokes its no-arg constructor, with the class name sourced from the manifest.properties entry of a model archive. The existing isAssignableFrom check correctly rejects classes that are not subtypes of the expected extension interface (BaseToolFactory for factory=, ArtifactSerializer for serializer-class-*), but the check runs after Class.forName() has already loaded and initialized the named class. Class.forName() with default initialization semantics executes the target class's static initializer before returning, so an attacker who can supply a crafted model archive can cause the static initializer of any class on the classpath to run during model loading, regardless of whether that class passes the subsequent type check. Exploitation requires a class with attacker-useful side effects in its static initializer (for example, JNDI lookup, outbound network I/O, or filesystem access) to be present on the classpath, so this is not a drop-in remote code execution; however, the attack surface grows as third-party model distribution becomes more common (community model repositories, Hugging Face-style sharing), where users routinely load model files from origins they do not control. A secondary, narrower vector affects deployments that ship legitimate BaseToolFactory or ArtifactSerializer subclasses with side-effecting no-arg constructors: a malicious manifest can name such a class and force its constructor to run during model load. Mitigation: * 2.x users should upgrade to 2.5.9. * 3.x users should upgrade to 3.0.0-M3. Note: The fix introduces a package-prefix allowlist that is consulted before Class.forName() is invoked, so the static initializer of a disallowed class is never executed. Classes under the opennlp. prefix remain permitted by default. Deployments that load models referencing factories or serializers outside opennlp.* must opt those packages in, either programmatically via ExtensionLoader.registerAllowedPackage(String) before the first model load, or by setting the OPENNLP_EXT_ALLOWED_PACKAGES system property to a comma-separated list of allowed package prefixes. Users who cannot upgrade immediately should ensure that all model files are sourced from trusted origins and should audit their classpath for classes with side-effecting static initializers or constructors, particularly any that perform JNDI lookups, network requests, or filesystem operations during class initialization.
Deserialization of Untrusted Data vulnerability in Apache IoTDB.This issue affects Apache IoTDB: from 0.13.0 through 0.13.4. Users are recommended to upgrade to version 1.2.2, which fixes the issue.
Apache James prior to version 3.7.5 and 3.8.0 exposes a JMX endpoint on localhost subject to pre-authentication deserialisation of untrusted data. Given a deserialisation gadjet, this could be leveraged as part of an exploit chain that could result in privilege escalation. Note that by default JMX endpoint is only bound locally. We recommend users to: - Upgrade to a non-vulnerable Apache James version - Run Apache James isolated from other processes (docker - dedicated virtual machine) - If possible turn off JMX
Apache OFBiz has unsafe deserialization prior to 17.12.07 version An unauthenticated user can perform an RCE attack
Hertzbeat is a real-time monitoring system. At the interface of `/define/yml`, SnakeYAML is used as a parser to parse yml content, but no security configuration is used, resulting in a YAML deserialization vulnerability. Version 1.4.1 fixes this vulnerability.
Deserialization of untrusted data in Apache Fory PyFory. PyFory's ReduceSerializer could bypass documented DeserializationPolicy validation hooks during reduce-state restoration and global-name resolution. An application is vulnerable if it deserializes attacker-controlled data using PyFory Python-native mode with strict mode disabled and relies on DeserializationPolicy to restrict unsafe classes, functions, or module attributes. This issue affects Apache Fory: from before 1.0.0. Mitigation: Users of Apache Fory are recommended to upgrade to version 1.0.0 or later, which enforces DeserializationPolicy validation for the affected ReduceSerializer paths and thus fixes this issue.
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
Deserialization of Untrusted Data vulnerability in Apache Camel PQC Component. The camel-pqc component persists post-quantum key metadata (KeyMetadata) through pluggable KeyLifecycleManager implementations. AwsSecretsManagerKeyLifecycleManager.deserializeMetadata() reads that metadata back from the configured AWS Secrets Manager secret by Base64-decoding the stored value and deserializing it with a raw java.io.ObjectInputStream.readObject() and no ObjectInputFilter or class allow-list; the cast to KeyMetadata happens only after readObject() returns, so any readObject() side effects in a crafted object run before the type check. A principal who can write to the AWS Secrets Manager secret that holds this metadata (requiring secretsmanager:PutSecretValue on that secret) could store a crafted serialized object that is deserialized during normal key-lifecycle operations, potentially leading to code execution in the context of the application that manages the keys. This is the same underlying defect, in the same code path and remediated by the same fix, as CVE-2026-46590, which was reported independently and additionally covers the HashiCorp Vault and file-based sibling managers; both are incomplete-remediation follow-ons to CVE-2026-40048 (CAMEL-23200). This issue affects Apache Camel: from 4.18.0 before 4.18.3, from 4.19.0 before 4.21.0. Users are recommended to upgrade to version 4.21.0, which fixes the issue. If users are on the 4.18.x LTS releases stream, then they are suggested to upgrade to 4.18.3. For deployments that cannot upgrade immediately, restrict write access to the AWS Secrets Manager secret that holds the camel-pqc key metadata so that only the application’s own identity holds secretsmanager:PutSecretValue on it (least-privilege IAM), and keep the PQC key material in a secret separate from any data that less-trusted principals can write.
The fix for CVE-2026-41409 was not applied to the 2.1.X and 2.2.X branches. Here was the original issue description: The fix for CVE-2024-52046 in Apache MINA AbstractIoBuffer.getObject() was incomplete. The classname allowlist of classes allowed to be deserialized was applied too late after a static initializer in a class to be read might already have been executed. Affected versions are Apache MINA 2.1.0 <= 2.1.11, and 2.2.0 <= 2.2.6. The problem is resolved in Apache MINA 2.1.12, and 2.2.7 by applying the classname allowlist earlier. Affected are applications using Apache MINA that call IoBuffer.getObject(). Applications using Apache MINA are advised to upgrade The fix for CVE-2024-52046 in Apache MINA AbstractIoBuffer.getObject() was incomplete. The classname allowlist of classes allowed to be deserialized was applied too late after a static initializer in a class to be read might already have been executed. Affected versions are Apache MINA 2.1.0 <= 2.1.110, and 2.2.0 <= 2.2.6. The problem is resolved in Apache MINA 2.1.12, and 2.2.7 by applying the classname allowlist earlier. Affected are applications using Apache MINA that call IoBuffer.getObject(). Applications using Apache MINA are advised to upgrade
Apache Dubbo is a java based, open source RPC framework. Versions prior to 2.6.10 and 2.7.10 are vulnerable to pre-auth remote code execution via arbitrary bean manipulation in the Telnet handler. The Dubbo main service port can be used to access a Telnet Handler which offers some basic methods to collect information about the providers and methods exposed by the service and it can even allow to shutdown the service. This endpoint is unprotected. Additionally, a provider method can be invoked using the `invoke` handler. This handler uses a safe version of FastJson to process the call arguments. However, the resulting list is later processed with `PojoUtils.realize` which can be used to instantiate arbitrary classes and invoke its setters. Even though FastJson is properly protected with a default blocklist, `PojoUtils.realize` is not, and an attacker can leverage that to achieve remote code execution. Versions 2.6.10 and 2.7.10 contain fixes for this issue.
A bug in `BaseSerialization.deserialize()` allowed unrestricted `import_string()` of attacker-controlled class paths when the Scheduler / API Server loaded a serialized DAG: a DAG author could embed a malicious trigger into a DAG to gain remote code execution on the API Server / Scheduler process, crossing the Airflow security boundary that DAG-author code must never execute in those processes. Users are advised to upgrade to `apache-airflow` 3.3.0 or later. As a defense-in-depth mitigation, deployments where DAG-author trust is limited can restrict the `[core] allowed_deserialization_classes` config to a narrow allowlist.
CWE-502 Deserialization of Untrusted Data at the rabbitmq-connector plugin module in Apache EventMesh (incubating) V1.7.0\V1.8.0 on windows\linux\mac os e.g. platforms allows attackers to send controlled message and remote code execute via rabbitmq messages. Users can use the code under the master branch in project repo to fix this issue, we will release the new version as soon as possible.
Deserialization of Untrusted Data vulnerability in Apache InLong.This issue affects Apache InLong: from 1.7.0 through 1.11.0, the attackers can bypass using malicious parameters. Users are advised to upgrade to Apache InLong's 1.12.0 or cherry-pick [1], [2] to solve it. [1] https://github.com/apache/inlong/pull/9694 [2] https://github.com/apache/inlong/pull/9707
XStream is a Java library to serialize objects to XML and back again. In XStream before version 1.4.16, there is a vulnerability which may allow a remote attacker to load and execute arbitrary code from a remote host only by manipulating the processed input stream. No user is affected, who followed the recommendation to setup XStream's security framework with a whitelist limited to the minimal required types. If you rely on XStream's default blacklist of the Security Framework, you will have to use at least version 1.4.16.
Unsafe deserialization occurs within a Dubbo application which has HTTP remoting enabled. An attacker may submit a POST request with a Java object in it to completely compromise a Provider instance of Apache Dubbo, if this instance enables HTTP. This issue affected Apache Dubbo 2.7.0 to 2.7.4, 2.6.0 to 2.6.7, and all 2.5.x versions.
Apache Tapestry 3.x allows deserialization of untrusted data, leading to remote code execution. This issue is similar to but distinct from CVE-2020-17531, which applies the the (also unsupported) 4.x version line. NOTE: This vulnerability only affects Apache Tapestry version line 3.x, which is no longer supported by the maintainer. Users are recommended to upgrade to a supported version line of Apache Tapestry.
Class org.apache.sshd.server.keyprovider.SimpleGeneratorHostKeyProvider in Apache MINA SSHD <= 2.9.1 uses Java deserialization to load a serialized java.security.PrivateKey. The class is one of several implementations that an implementor using Apache MINA SSHD can choose for loading the host keys of an SSH server.
A flaw was discovered in jackson-databind in versions before 2.9.10, 2.8.11.5 and 2.6.7.3, where it would permit polymorphic deserialization of a malicious object using commons-configuration 1 and 2 JNDI classes. An attacker could use this flaw to execute arbitrary code.
Apache Jena SDB 3.17.0 and earlier is vulnerable to a JDBC Deserialisation attack if the attacker is able to control the JDBC URL used or cause the underlying database server to return malicious data. The mySQL JDBC driver in particular is known to be vulnerable to this class of attack. As a result an application using Apache Jena SDB can be subject to RCE when connected to a malicious database server. Apache Jena SDB has been EOL since December 2020 and users should migrate to alternative options e.g. Apache Jena TDB 2.
A deserialization vulnerability existed in dubbo hessian-lite 3.2.12 and its earlier versions, which could lead to malicious code execution. This issue affects Apache Dubbo 2.7.x version 2.7.17 and prior versions; Apache Dubbo 3.0.x version 3.0.11 and prior versions; Apache Dubbo 3.1.x version 3.1.0 and prior versions.
Deserialization of untrusted data in IPC and Parquet readers in PyArrow versions 0.14.0 to 14.0.0 allows arbitrary code execution. An application is vulnerable if it reads Arrow IPC, Feather or Parquet data from untrusted sources (for example user-supplied input files). This vulnerability only affects PyArrow, not other Apache Arrow implementations or bindings. It is recommended that users of PyArrow upgrade to 14.0.1. Similarly, it is recommended that downstream libraries upgrade their dependency requirements to PyArrow 14.0.1 or later. PyPI packages are already available, and we hope that conda-forge packages will be available soon. If it is not possible to upgrade, we provide a separate package `pyarrow-hotfix` that disables the vulnerability on older PyArrow versions. See https://pypi.org/project/pyarrow-hotfix/ for instructions.
The Java OpenWire protocol marshaller is vulnerable to Remote Code Execution. This vulnerability may allow a remote attacker with network access to either a Java-based OpenWire broker or client to run arbitrary shell commands by manipulating serialized class types in the OpenWire protocol to cause either the client or the broker (respectively) to instantiate any class on the classpath. Users are recommended to upgrade both brokers and clients to version 5.15.16, 5.16.7, 5.17.6, or 5.18.3 which fixes this issue.
Apache Geode versions up to 1.12.5, 1.13.4 and 1.14.0 are vulnerable to a deserialization of untrusted data flaw when using JMX over RMI on Java 8. Any user still on Java 8 who wishes to protect against deserialization attacks involving JMX or RMI should upgrade to Apache Geode 1.15 and Java 11. If upgrading to Java 11 is not possible, then upgrade to Apache Geode 1.15 and specify "--J=-Dgeode.enableGlobalSerialFilter=true" when starting any Locators or Servers. Follow the documentation for details on specifying any user classes that may be serialized/deserialized with the "serializable-object-filter" configuration option. Using a global serial filter will impact performance.
Deserialization of Untrusted Data vulnerability in Apache Dubbo.This issue only affects Apache Dubbo 3.1.5. Users are recommended to upgrade to the latest version, which fixes the issue.
Java object deserialization issue in Jackrabbit webapp/standalone on all platforms allows attacker to remotely execute code via RMIVersions up to (including) 2.20.10 (stable branch) and 2.21.17 (unstable branch) use the component "commons-beanutils", which contains a class that can be used for remote code execution over RMI. Users are advised to immediately update to versions 2.20.11 or 2.21.18. Note that earlier stable branches (1.0.x .. 2.18.x) have been EOLd already and do not receive updates anymore. In general, RMI support can expose vulnerabilities by the mere presence of an exploitable class on the classpath. Even if Jackrabbit itself does not contain any code known to be exploitable anymore, adding other components to your server can expose the same type of problem. We therefore recommend to disable RMI access altogether (see further below), and will discuss deprecating RMI support in future Jackrabbit releases. How to check whether RMI support is enabledRMI support can be over an RMI-specific TCP port, and over an HTTP binding. Both are by default enabled in Jackrabbit webapp/standalone. The native RMI protocol by default uses port 1099. To check whether it is enabled, tools like "netstat" can be used to check. RMI-over-HTTP in Jackrabbit by default uses the path "/rmi". So when running standalone on port 8080, check whether an HTTP GET request on localhost:8080/rmi returns 404 (not enabled) or 200 (enabled). Note that the HTTP path may be different when the webapp is deployed in a container as non-root context, in which case the prefix is under the user's control. Turning off RMIFind web.xml (either in JAR/WAR file or in unpacked web application folder), and remove the declaration and the mapping definition for the RemoteBindingServlet: <servlet> <servlet-name>RMI</servlet-name> <servlet-class>org.apache.jackrabbit.servlet.remote.RemoteBindingServlet</servlet-class> </servlet> <servlet-mapping> <servlet-name>RMI</servlet-name> <url-pattern>/rmi</url-pattern> </servlet-mapping> Find the bootstrap.properties file (in $REPOSITORY_HOME), and set rmi.enabled=false and also remove rmi.host rmi.port rmi.url-pattern If there is no file named bootstrap.properties in $REPOSITORY_HOME, it is located somewhere in the classpath. In this case, place a copy in $REPOSITORY_HOME and modify it as explained.
An attacker can use SnakeYAML to deserialize java.net.URLClassLoader and make it load a JAR from a specified URL, and then deserialize javax.script.ScriptEngineManager to load code using that ClassLoader. This unbounded deserialization can likely lead to remote code execution. The code can be run in Helix REST start and Workflow creation. Affect all the versions lower and include 1.2.0. Affected products: helix-core, helix-rest Mitigation: Short term, stop using any YAML based configuration and workflow creation. Long term, all Helix version bumping up to 1.3.0
A Deserialization of Untrusted Data vulnerability exists in the Apache ActiveMQ NMS AMQP Client. This issue affects all versions of Apache ActiveMQ NMS AMQP up to and including 2.3.0, when establishing connections to untrusted AMQP servers. Malicious servers could exploit unbounded deserialization logic present in the client to craft responses that may lead to arbitrary code execution on the client side. Although version 2.1.0 introduced a mechanism to restrict deserialization via allow/deny lists, the protection was found to be bypassable under certain conditions. In line with Microsoft’s deprecation of binary serialization in .NET 9, the project is evaluating the removal of .NET binary serialization support from the NMS API entirely in future releases. Mitigation and Recommendations: Users are strongly encouraged to upgrade to version 2.4.0 or later, which resolves the issue. Additionally, projects depending on NMS-AMQP should migrate away from .NET binary serialization as part of a long-term hardening strategy.
Deserialization of Untrusted Data vulnerability in Apache Seata (incubating). This issue affects Apache Seata (incubating): 2.4.0. Users are recommended to upgrade to version 2.5.0, which fixes the issue.
The Solr plugin of Apache OFBiz is configured by default to automatically make a RMI request on localhost, port 1099. In version 18.12.05 and earlier, by hosting a malicious RMI server on localhost, an attacker may exploit this behavior, at server start-up or on a server restart, in order to run arbitrary code. Upgrade to at least 18.12.06 or apply patches at https://issues.apache.org/jira/browse/OFBIZ-12646.
The REST Plugin in Apache Struts 2.1.1 through 2.3.x before 2.3.34 and 2.5.x before 2.5.13 uses an XStreamHandler with an instance of XStream for deserialization without any type filtering, which can lead to Remote Code Execution when deserializing XML payloads.
Deserialization of Untrusted Data vulnerability in Apache Software Foundation Apache InLong.This issue affects Apache InLong: from 1.1.0 through 1.5.0. Users are advised to upgrade to Apache InLong's latest version or cherry-pick https://github.com/apache/inlong/pull/7223 https://github.com/apache/inlong/pull/7223 to solve it.
A deserialization vulnerability existed when dubbo generic invoke, which could lead to malicious code execution. This issue affects Apache Dubbo 2.7.x version 2.7.21 and prior versions; Apache Dubbo 3.0.x version 3.0.13 and prior versions; Apache Dubbo 3.1.x version 3.1.5 and prior versions.
In Apache Linkis <=1.3.1, because the parameters are not effectively filtered, the attacker uses the MySQL data source and malicious parameters to configure a new data source to trigger a deserialization vulnerability, eventually leading to remote code execution. Versions of Apache Linkis <= 1.3.0 will be affected. We recommend users upgrade the version of Linkis to version 1.3.2.
In Apache Linkis <=1.3.1, due to the lack of effective filtering of parameters, an attacker configuring malicious Mysql JDBC parameters in JDBC EengineConn Module will trigger a deserialization vulnerability and eventually lead to remote code execution. Therefore, the parameters in the Mysql JDBC URL should be blacklisted. Versions of Apache Linkis <= 1.3.0 will be affected. We recommend users upgrade the version of Linkis to version 1.3.2.
A deserialization vulnerability existed when decode a malicious package.This issue affects Apache Dubbo: from 3.1.0 through 3.1.10, from 3.2.0 through 3.2.4. Users are recommended to upgrade to the latest version, which fixes the issue.
Deserialization of Untrusted Data vulnerability in Apache Seata (incubating). This security vulnerability is the same as CVE-2024-47552, but the version range described in the CVE-2024-47552 definition is too narrow. This issue affects Apache Seata (incubating): from 2.0.0 before 2.3.0. Severity Justification: The Apache Seata security team assesses the severity of this vulnerability as "Low" due to stringent real-world mitigating factors. First, the vulnerability is strictly isolated to the Raft cluster mode, an optional and non-default feature introduced in v2.0.0, while most users rely on the unaffected traditional architecture. Second, Seata is an internal middleware; communication between TC and RM/TM occurs entirely within trusted internal networks. An attacker would require prior, unauthorized access to the Intranet to exploit this, making external exploitation highly improbable. Users are recommended to upgrade to version 2.3.0, which fixes the issue.
Deserialization of Untrusted Data vulnerability in Apache ActiveMQ NMS OpenWire Client. This issue affects Apache ActiveMQ NMS OpenWire Client before 2.1.1 when performing connections to untrusted servers. Such servers could abuse the unbounded deserialization in the client to provide malicious responses that may eventually cause arbitrary code execution on the client. Version 2.1.0 introduced a allow/denylist feature to restrict deserialization, but this feature could be bypassed. The .NET team has deprecated the built-in .NET binary serialization feature starting with .NET 9 and suggests migrating away from binary serialization. The project is considering to follow suit and drop this part of the NMS API altogether. Users are recommended to upgrade to version 2.1.1, which fixes the issue. We also recommend to migrate away from relying on .NET binary serialization as a hardening method for the future.