Deserialization Vulnerability in NVIDIA Megatron Bridge Could Enable Remote Code Execution

NVIDIA Megatron Bridge contains a vulnerability where an attacker could cause a deserialization of untrusted data. A successful exploit of this vulnerability might lead to code execution, data tampering, and information disclosure.

Insecure Deserialization in NVIDIA Megatron Bridge Enables Remote Code Execution

NVIDIA Megatron Bridge contains a vulnerability where an attacker could cause a deserialization of untrusted data. A successful exploit of this vulnerability might lead to code execution, data tampering, and information disclosure.

ModelScope through 1.40.0 Unsafe YAML Deserialization in Model Config Loading

ModelScope uses PyYAML's unsafe yaml.Loader to parse model configuration files, allowing arbitrary code execution through Python object construction tags. Attackers can craft malicious model repositories with poisoned configuration files that execute code when loaded by users.

Unsafe deserialization of decrypted terms enables node DoS in AshCloak

Deserialization of Untrusted Data vulnerability in ash-project ash_cloak allows an attacker who can influence the bytes of an encrypted column to crash the BEAM node, by triggering unbounded atom creation or a decompression bomb during decryption. AshCloak.Calculations.Decrypt decodes the decrypted binary with Ash.Helpers.non_executable_binary_to_term/1 without the :safe option, so atoms in the payload are interned during the decode and never garbage collected, and the term format's compressed form is inflated transparently. vault.decrypt!() is the only barrier and stops tampering only for an authenticated cipher. Cloak also ships the unauthenticated AES.CTR, whose ciphertext an attacker who knows their own plaintext can XOR into any same-length payload without the key, so an ordinary read of the forged column reaches the decoder. A few hundred kilobytes of distinct atoms exhausts the atom table, or a small compressed payload inflates to gigabytes. This issue affects ash_cloak: from 0.1.0 before 0.4.0.

Remote JVM termination: nested-array Java deserialization bypasses allowlist, triggers StackOverflowError, default JavaLangErrorHandler calls System.exit(99)

A single hostile AMQP message can terminate the entire consumer JVM (System.exit(99)), not just the listener thread — full availability loss for every workload co-located in that process. Spring AMQP 4.1.0 Spring AMQP 4.0.0 - 4.0.4 Spring AMQP 3.2.0 - 3.2.12 Spring AMQP 2.4.18 and earlier

jackson-databind omits java.lang.Comparable from DefaultBaseTypeLimitingValidator's unsafe base types

DefaultBaseTypeLimitingValidator is the PolymorphicTypeValidator applied automatically whenever @JsonTypeInfo is used without an explicitly configured custom validator. It denies polymorphic resolution only for a fixed set of "unsafe base types", and its isSafeSubType method returns true unconditionally for every base type outside that set. java.lang.Comparable was absent from the list despite being implemented by a very large fraction of JDK and application classes, comparable in breadth to java.io.Serializable, which is on the list for that reason. An application declaring an @JsonTypeInfo-annotated property or class with Comparable as its base type, and no custom PolymorphicTypeValidator, will accept a type identifier for essentially any class implementing Comparable. This yields an attacker-controlled object instantiation primitive; a demonstrated case constructs a java.io.File for an arbitrary attacker-chosen path, which becomes path-traversal-adjacent if the application subsequently calls path-sensitive methods on the value. No class implementing Comparable has been identified that yields code execution through deserialization alone. Global Default Typing via activateDefaultTyping is not affected, because that method structurally requires an explicit PolymorphicTypeValidator argument. This affects com.fasterxml.jackson.core:jackson-databind from 2.11.0 before 2.18.10, from 2.19.0 before 2.21.6, and from 2.22.0 before 2.22.2, and tools.jackson.core:jackson-databind from 3.0.0 before 3.1.6 and from 3.2.0 before 3.2.2. Users should upgrade to 2.18.10, 2.21.6, 2.22.2, 3.1.6, or 3.2.2.

Jackson Configuration Allows Code Execution with Unknown "Serialization Gadgets"

When configured to enable default typing, Jackson contained a deserialization vulnerability that could lead to arbitrary code execution. Jackson fixed this vulnerability by blacklisting known "deserialization gadgets". Spring Batch configures Jackson with global default typing enabled which means that through the previous exploit, arbitrary code could be executed if all of the following is true: * Spring Batch's Jackson support is being leveraged to serialize a job's ExecutionContext. * A malicious user gains write access to the data store used by the JobRepository (where the data to be deserialized is stored). In order to protect against this type of attack, Jackson prevents a set of untrusted gadget classes from being deserialized. Spring Batch should be proactive against blocking unknown "deserialization gadgets" when enabling default typing.

WordPress Tickera plugin <= 3.6.0.2 - PHP Object Injection vulnerability

Unauthenticated PHP Object Injection in Tickera <= 3.6.0.2 versions.

jackson-databind resolves attacker-controlled URI schemes when deserializing java.nio.file.Path

jackson-databind's deserializer for java.nio.file.Path resolves an attacker-supplied URI without restricting the URI scheme. In JDKFromStringDeserializer.NioPathHelper.deserialize, a string bound from untrusted JSON is passed to new URI(value) and then to Path.of(uri). When that throws FileSystemNotFoundException, the code enumerates ServiceLoader<FileSystemProvider> and calls provider.getPath(uri) on the first provider whose scheme matches the attacker-chosen scheme. Untrusted JSON can therefore select and drive an arbitrary registered FileSystemProvider during readValue under a default JsonMapper, and forces provider class loading at the same time. With only the JDK built-in providers (file, jar/zipfs) present, the resolved path is inert and no mount or network I/O occurs; further impact requires a side-effecting third-party FileSystemProvider on the classpath. This affects com.fasterxml.jackson.core:jackson-databind from 2.8.0 before 2.18.10, from 2.19.0 before 2.21.6, and from 2.22.0 before 2.22.2, and tools.jackson.core:jackson-databind from 3.0.0 before 3.1.6 and from 3.2.0 before 3.2.2. Users should upgrade to 2.18.10, 2.21.6, 2.22.2, 3.1.6, or 3.2.2. Binding java.nio.file.Path from untrusted JSON should be avoided regardless of version.

Unrestricted Java Deserialization in OpenSearch SQL Plugin Cursor Pagination

Unrestricted deserialization of untrusted data in the cursor pagination component in the OpenSearch SQL plugin allows a remote authenticated user with basic read/search permissions to execute arbitrary code on the server by sending a crafted cursor parameter to the plugins/sql endpoint.

vLLM: OpenAI auth bypass

vLLM is an inference and serving engine for large language models (LLMs). From 0.3.0 until 0.22.0, a vulnerability in ASGI web servers and starlette's trust on those web servers enables an authentication bypass of the OpenAI API AuthenticationMiddleware. It allows to use the API without providing the configured VLLM_API_KEY or --api-key. This vulnerability is fixed in 0.22.0.

An issue in Beijing Tongtech Co., Ltd tongweb v.7.0.24 in the Spring HttpInovkerServiceExporter component allows a remote attacker to execute arbitrary code via a crafted request to the console/heimdall endpoint

Pimcore Hotspotimage getDataFromResource() unrestricted Serialize::unserialize over object-store column

Pimcore is an Open Source Data & Experience Management Platform. Prior to 11.5.19, 12.3.10, and 2026.1.6, Pimcore\Model\DataObject\ClassDefinition\Data\Hotspotimage::getDataFromResource() in models/DataObject/ClassDefinition/Data/Hotspotimage.php passes the field __hotspots object-store column to Pimcore\Tool\Serialize::unserialize() without an allowed-classes restriction after JSON decoding fails. An attacker with a separate capability to write crafted PHP serialized bytes into that column can instantiate available classes and trigger magic methods when an affected DataObject is loaded, which can produce arbitrary file writes or code execution through bundled gadget chains. The related ImageGallery, Block, and Video callers use the same fallback pattern, but the identified June fix changes the Hotspotimage caller only. This issue is fixed for Hotspotimage in versions 11.5.19, 12.3.10, and 2026.1.6.

Deserialization allow-list silently bypassed: setBeanClassLoader replaces deserializer but mapper keeps stale reference

An operator who calls JdbcMessageStore.addAllowedPatterns(...) to restrict deserialization receives no protection at all when the store is a Spring-managed bean. Spring Integration 7.1.0 Spring Integration 7.0.0 - 7.0.5 Spring Integration 6.5.0 - 6.5.10 Spring Integration 6.4.0 - 6.4.12

SvelteKit 2.49.0 before 2.53.3 Denial of Service via form

SvelteKit versions from 2.49.0 through 2.53.2 (fixed in 2.53.3) contain a deserialization expansion issue in the experimental form remote function. When an application enables experimental.remoteFunctions and uses the form function to process the files array without validating files.length or individual file sizes, an attacker can submit relatively small inputs that expand into very large file arrays, leading to expensive processing and denial of service.

Remote Code Execution in extension "HTML5 Video Player vs. Powermail" (html5videoplayer_powermail)

The extension fails to safely process untrusted client input of an attacker-controlled cookie directly to PHP's unserialize(). A remote, unauthenticated attacker can supply a crafted serialized payload to trigger PHP Object Injection, leading to Remote Code Execution on the TYPO3 server.

Insecure Deserialization in extension "Apache Solr for TYPO3 - Enterprise Search" (solr)

The extension's indexer passed every field value returned by content object rendering through PHP's unserialize() function when transferring multi-value data for the SOLR_CLASSIFICATION, SOLR_MULTIVALUE and SOLR_RELATION content object types, rather than a safe format. If user-generated content saved in the TYPO3 database can reach an indexed field, this exposes a PHP Object Injection surface.

WordPress Geo Controller plugin <= 8.9.8 - PHP Object Injection vulnerability

Unauthenticated PHP Object Injection in Geo Controller <= 8.9.8 versions.

SOY CMS contains an issue with deserialization of untrusted data. An arbitrary code may be executed by an attacker with the web server privilege.

SpeechBrain < 1.1.1 Arbitrary Code Execution via CKPT.yaml Parsing

SpeechBrain before 1.1.1 contains an arbitrary code execution vulnerability that allows attackers to execute arbitrary code by supplying a crafted CKPT.yaml checkpoint metadata file parsed with PyYAML's unsafe loader during candidate enumeration in Checkpointer.recover_if_possible(). Attackers can embed malicious Python object construction tags such as !!python/object/apply in any CKPT.yaml file within the configured checkpoint path to trigger code execution during candidate discovery, even if the malicious checkpoint is never selected for recovery.