appsec.fyi

Insecure Deserialization Resources

Post Share

A curated AppSec resource library covering XSS, SQLi, SSRF, IDOR, RCE, XXE, OSINT, and more.

Insecure Deserialization

Insecure deserialization vulnerabilities occur when applications reconstruct objects from serialized data without proper validation, potentially allowing attackers to execute arbitrary code, bypass authentication, or manipulate application logic. These vulnerabilities affect virtually every major programming language: Java (ObjectInputStream, ysoserial gadget chains), PHP (unserialize), Python (pickle, PyYAML), .NET (BinaryFormatter, Json.NET with TypeNameHandling), and Ruby (Marshal). Deserialization attacks are particularly dangerous because they often achieve remote code execution with a single crafted payload. The exploitation landscape includes gadget chain discovery, polyglot payloads that work across libraries, and attacks against message queues, caching layers, and session management systems that serialize user-controlled data. Defenses include avoiding native serialization for untrusted data, using safe alternatives like JSON, implementing allowlists for deserialized types, and integrity checking serialized objects.

The vulnerability is the type information, not the parser

Insecure deserialization is regularly explained as a parsing problem, which makes it sound like input validation would help. It would not. The issue is that serialization formats such as Java's native serialization, .NET's BinaryFormatter, PHP's unserialize and Python's pickle encode which classes to instantiate alongside the data. Reconstructing the object therefore means running that type's reconstruction logic, and in every one of these runtimes there is a hook that executes during reconstruction — readObject, __wakeup and __destruct, __reduce__. The attacker is not injecting code. They are choosing which of your existing code runs, and in what order.

That is what a gadget chain is: a sequence of ordinary, non-malicious classes already present in the application or its dependencies, arranged so that the side effects of deserializing them compose into something useful. This is why the vulnerability depends so heavily on what is on the classpath, why tools like ysoserial ship a catalogue of chains per library, and why adding a dependency can make a previously unexploitable endpoint exploitable without any change to your own code.

It is also why blocklists fail. Forbidding the classes in a published gadget chain prevents that chain and not the next one, and the history of Java deserialization mitigations is a long sequence of new chains found in libraries nobody had examined. Allowlisting the specific types an endpoint legitimately expects is the only version of type filtering that holds, and it requires knowing that set.

The durable fix is to not accept serialized objects from untrusted sources at all. A data-only format — JSON, or a schema-bound format like Protocol Buffers — carries values rather than types, so there is nothing to instantiate. Where an object graph genuinely must cross a trust boundary, signing it and verifying before deserializing moves the problem to key management, which is a better problem.

Look for these wherever serialized data travels: cookies and hidden fields, caches, message queues, session stores and inter-service calls.

Start here: the Deser learning pathAn ordered route through this library — orientation, then methodology, then research. Read the Deser guideA long-form, source-cited deep dive synthesized from every resource below. The comprehensive Deser guide on chs.usA hand-written, in-depth practitioner guide — attacks, testing, and prevention.
Date Added Link Excerpt
2026-11-20 NEW 2026WP STAGING PHP Object Injection newsLibrary advisory for GHSA-4v8m-v3r8-w9q4 detailing a PHP Object Injection vulnerability in WP STAGING and WP STAGING Pro plugins (versions up to 4.17.0 and 6.17.0). Exploitation, achievable by authenticated users with Subscriber-level access, leverages deserialization of untrusted input during database restore or staging reset. Attackers can inject a malicious serialized object, leading to arbitrary code execution via a built-in POP chain when an administrator performs a restore or reset. Additional gadget chains facilitate arbitrary local file reads and, in WP STAGING Pro, arbitrary file deletion.
2026-11-20 NEW 2026Insecure deserialization as a DAST finding: confirming it properly intermediate 24 min readLibrary for confirming insecure deserialization findings from dynamic scans, detailing scenarios involving Python's `pickle` and PHP's `unserialize()`. It highlights how scanners can misinterpret findings, such as mistaking a sleep handler for time-based deserialization or failing to detect active code execution behind a false negative. The resource analyzes the impact of deserialization through gadget chains, magic methods like `__wakeup` and `__destruct`, and the role of `allowed_classes` in PHP, emphasizing that the vulnerability lies in deserializing untrusted data with native object formats.
2026-11-20 NEW 2026NVIDIA Megatron-LM checkpoint loading deserialization news AI PythonNVIDIA Megatron-LM has a vulnerability related to checkpoint loading and deserialization. This flaw allows for arbitrary code execution if a malicious checkpoint file is loaded. The exploit poses a significant security risk. Users are advised to update to the latest versions of Megatron-LM to mitigate this vulnerability.
2026-11-20 NEW 2026NLTK AllowlistUnpickler doesn't deny numpy file-I/O sinks news 4 min read PythonLibrary for Python's NLTK that prevents arbitrary file reads and constrained file creations via crafted pickle files. Versions nltk >= 3.10.0 and <= 3.10.3 were vulnerable because `allowlisted_pickle_load` incorrectly keyed its denylist by export name, allowing access to `numpy.lib._npyio_impl.recfromtxt` for file reads and `numpy.lib.format.open_memmap` for file creation when NumPy was installed. The fix involves denying entire `numpy.lib` prefixes and using an exact `_MODEL_ALLOWED_GLOBALS` list, with the vulnerability re-assessed to Medium severity.
2026-11-19 NEW 2026Deserialization of untrusted data in cookie/form validation advancedThis vulnerability arises when an application deserializes untrusted data within the double-submit cookie mechanism used for CSRF protection or form validation. If an attacker can control the serialized data sent to the application, they can exploit this flaw to achieve remote code execution. The attack involves crafting malicious serialized objects that, when deserialized, trigger unintended code execution on the server, potentially leading to a full system compromise. This highlights the critical need to sanitize or avoid deserializing untrusted input.
2026-11-19 NEW 2026Oracle Platform Security for Java critical deserialization intermediateLibrary with critical deserialization vulnerability in Oracle Platform Security for Java (versions 12.2.1.4.0 and 14.1.2.0.0). This easily exploitable flaw, identified as GHSA-qc49-p4c3-3hxc, allows unauthenticated attackers with network access via HTTP to achieve a complete takeover of the product, impacting confidentiality, integrity, and availability with a CVSS 3.1 Base Score of 10.0.
2026-11-19 NEW 2026Masteriyo LMS WordPress PHP object injection intermediateWriteup of GHSA-5xv5-3557-x7h5, a PHP Object Injection vulnerability in Masteriyo LMS versions up to 3.5.1. Attackers can exploit this flaw by deserializing untrusted input via the course importer's `set_post_metas()` method. While unauthenticated exploitation is possible under specific configurations (enable_instructor_registration and enable_guest_checkout enabled, plus a purchasable course), it requires author-level access on default installations.
2026-11-19 NEW 2026WPvivid Backup deserialization intermediateWPvivid Backup, a WordPress plugin, has a critical deserialization vulnerability that allows unauthenticated attackers to execute arbitrary code on a WordPress site. This issue arises from improper handling of serialized data within the plugin's functionality. Attackers can exploit this vulnerability to gain full control of a website. Users are strongly advised to update WPvivid Backup to the latest version to patch this security flaw.
2026-11-19 NEW 2026NVIDIA Model Optimizer unsafe deserialization intermediate AINVIDIA Model Optimizer for Linux and Windows has a critical vulnerability due to unsafe deserialization. This flaw allows remote attackers to execute arbitrary code on affected systems by sending specially crafted serialized data. The vulnerability is present in versions prior to 1.0.0. NVIDIA has released an update to address this issue. Users are strongly advised to update to the latest version to mitigate the security risk.
2026-11-18 NEW 2026Java4Shell: RCE in OpenJDK when processing Base64 news 14 min read RCEWriteup on Java4Shell (CVE-2026-114938), a critical memory-corruption vulnerability in OpenJDK's Base64 decoder. Exploitation allows for heap memory overwrite, data leakage akin to Heartbleed, and remote code execution. The vulnerability is present in widely deployed Java applications including Keycloak, Jira, Confluence, Jenkins, and Kafka, affecting default configurations and potentially others utilizing features like authentication and file uploads. The discovery highlights the risks associated with optimizations in shared software stacks and advocates for "vibe-uncoding" by removing unnecessary code to reduce attack surface.
2026-11-18 NEW 2026CVE-2026-114938 RCE in OpenJDK: Everything You Need to Know news 3 min read RCEWriteup of CVE-2026-114938, "Java4Shell," detailing a critical OpenJDK RCE vulnerability. This flaw stems from a heap memory corruption during AVX-512 Base64 decoding on x86-64 systems, impacting various popular Java applications like Jenkins, Confluence, and Apache Tomcat. Exploitation is feasible, even with AI assistance, and patching to updated OpenJDK versions or applying JVM workarounds is recommended for affected distributions and hardware. → wiz.io
2026-11-18 NEW 2026Java4Shell: pre-auth RCE flaw in Java 17, 21, 25 and 27, with affected versions and fix (CVE-2026-114938) news RCEWriteup of CVE-2026-114938, a pre-auth RCE memory corruption flaw in OpenJDK's Base64 decoder affecting Java 17, 21, 25, and 27. Exploitation requires an affected Java build on x86-64 CPUs with AVX-512 VBMI, and attacker-controlled data reaching the decoder. Oracle and Eclipse Temurin have released patches, with Aikido offering patched Java images for Temurin 17 and 21. → aikido.dev
2026-11-18 NEW 2026Puppet Enterprise CVE-2026-115731 deserialization RCE news 7 min read RCELibrary for analyzing CVE-2026-115731, a deserialization RCE vulnerability in Puppet Enterprise and OpenVox Server. The bug allows arbitrary code execution as the puppet user by sending a crafted agent report to the `/puppet/v3/report` endpoint, potentially leading to CA signing key compromise and fleet-wide impersonation. If autosign is enabled, this vulnerability becomes unauthenticated RCE. The writeup details the exploitation path via `Types::TypeParser` and `RubyGenerator`, and provides remediation steps including patching, disabling autosign, and hunting for compromise indicators.
2026-11-18 NEW 2026AWS advanced JDBC wrapper: Java deserialization RCE via cache poisoning intermediate RCELibrary for AWS Advanced JDBC Wrapper addresses CVE-2026-14265, a Java deserialization RCE vulnerability. When the `RemoteQueryCachePlugin` is active, it deserializes cached query results without proper class filtering. An attacker with write access to the Redis/Valkey cache can poison it with a malicious serialized object, leading to arbitrary code execution on the application server. Disabling the plugin or restricting cache write access mitigates this.
2026-11-18 NEW 2026RBH Integra32: unauthenticated .NET Remoting deserialization intermediate 1 min read RCEWriteup detailing an unauthenticated deserialization vulnerability in RBH Integra32's .NET Remoting service. The flaw, exploitable via TCP port 2001 using tools like `ysoserial.exe` and `ExploitRemotingService.exe`, allows for arbitrary file read/write and blind Remote Code Execution as NT AUTHORITY\SYSTEM by overwriting critical system files or by leveraging gadget chains like `TypeConfuseDelegate`.
2026-11-18 NEW 2026InIt CTF: Super Pickle intermediate 3 min read PythonWriteup detailing the exploitation of Python's `pickle` module in the InIt CTF's "Super Pickle" challenge. The challenge involved deserializing attacker-controlled pickle data, which, due to the `REDUCE` opcode and the `__reduce__` method, allows arbitrary function execution. Despite keyword filters targeting common RCE gadgets like `system` and `open`, the writeup demonstrates successful exploitation using lesser-known standard library functions such as `pty.spawn`, `glob.glob`, and `linecache.getlines` to achieve code execution and subsequently read flag-related environment variables.
2026-11-17 NEW 2026Testimonial Slider deserialization of untrusted data intermediateA deserialization vulnerability in the Testimonial Slider plugin allows for arbitrary code execution. Attackers can exploit this by crafting a malicious serialized object and uploading it to the plugin's settings. This could lead to the compromise of entire WordPress sites.
2026-11-17 NEW 2026Optimole WordPress plugin deserialization intermediateAdvisory for CVE-2026-124969 details a vulnerability in the Optimole WordPress plugin (versions up to 4.2.19) allowing arbitrary file deletion. This occurs due to insufficient path validation in `replace_urls_in_value`, potentially leading to remote code execution if PEAR's Archive_Tar library is present and an administrator renames an attachment after a malicious payload is staged in a comment.
2026-11-17 NEW 2026httpx Ruby gem unsafe JSON deserialization intermediateLibrary advisory for GHSA-jgqw-w3x2-9h6q impacting the httpx Ruby gem. Versions prior to 1.9.1 are vulnerable to unsafe JSON deserialization when the oj gem is used and multi_json is absent. This vulnerability allows attackers to control object instantiation and potentially achieve remote code execution by influencing the JSON response body of fetched URLs. The issue stems from `Oj.load` using default `:object` mode, enabling class allocation and instance variable manipulation via specific JSON structures.
2026-11-17 NEW 20263DEXPERIENCE Station Launcher deserialization (critical) intermediate RCEA critical deserialization vulnerability exists in the 3DEXPERIENCE Station Launcher, allowing unauthenticated remote attackers to execute arbitrary code. This flaw stems from the improper handling of untrusted data. Successful exploitation could lead to a complete system compromise.
2026-11-17 NEW 2026SharePoint deserialization of untrusted data intermediate RCEA vulnerability in Microsoft Office allows for deserialization of untrusted data in SharePoint. This means that specially crafted data could be processed, potentially leading to unauthorized code execution on the server. Attackers could exploit this by tricking users into interacting with malicious content or by manipulating data within SharePoint itself. Successful exploitation could compromise the entire SharePoint environment.
2026-11-16 NEW 2026Document Library Pro deserialization of untrusted data intermediateA deserialization of untrusted data vulnerability has been identified in Barn2's Document Library Pro plugin. This flaw allows attackers to execute arbitrary code on the server by manipulating serialized data. Successful exploitation could lead to a complete compromise of the website. Users are strongly advised to update to the latest version of Document Library Pro to patch this critical security issue.
2026-11-16 NEW 2026JGD: an autonomous agent for Java deserialization gadget chains intermediate 3 min read AITool automating Java deserialization research, JGD identifies and generates exploit gadget chains and weaponized proof-of-concept payloads from JAR directories. It supports discovering both known and novel chains, including those within the Vavr, Spring AOP, Guava, and other library families, and outputs serialized payloads with RCE-closure demonstrations. JGD utilizes static and dynamic analysis, adversarial LLM auditing, and a comprehensive database of public chains to automate the end-to-end process of finding and weaponizing deserialization vulnerabilities.
2026-11-16 NEW 2026AI broke the deserialization threat model advanced 8 min read AILibrary for dynamic runtime protection against AI-accelerated Java deserialization attacks. It addresses the critical vulnerability where AI tools like GadgetHunter and LLM-based agents rapidly discover novel gadget chains, bypassing traditional signature-based defenses and vendor denylists. This library confines deserialization operations to restricted micro-compartments, blocking malicious behavior without relying on prior knowledge of exploits such as those found against WildFly, WebSphere, and Payara servers.
2026-11-16 NEW 2026247CTF Meme Upload Service — Hand-Built Phar Meets XXE intermediate 5 min read RCE XXEWriteup detailing a chained XXE and PHAR deserialization vulnerability in a PHP 7.3 application. The attack leverages `DOMDocument`'s `LIBXML_DTDLOAD` option to resolve external DTDs via `phar://` streams, triggering PHAR metadata deserialization in older PHP versions. This process allows for the injection of a serialized `Message` object, whose destructor writes a web shell to the document root. The article emphasizes overcoming silent failures by using observable side effects as oracles for validating crafted PHAR structures within byte limits.
2026-11-15 NEW 2026BSON serialization deserialization type-restriction bypass intermediateLibrary for BSON type-restriction bypass vulnerabilities, enabling unauthorized construction and population of unintended object types within consuming applications. This bypass exploits an overly permissive heuristic in class type checking, allowing attackers with BSON content insertion capabilities to trigger unexpected object states and downstream logic errors without requiring elevated privileges.
2026-11-15 NEW 2026Edito CMS unsafe JSON deserialization intermediateEdito CMS has a critical security vulnerability due to unsafe JSON deserialization. This flaw allows attackers to exploit the system by crafting malicious JSON payloads. Successful exploitation could lead to remote code execution, enabling attackers to compromise the entire system. Users are strongly advised to update to the latest version of Edito CMS to patch this vulnerability.
2026-11-15 NEW 2026OpenEMR 8.3.0 unsanitized deserialization intermediate RCEOpenEMR versions prior to 8.3.0 are vulnerable to unsanitized deserialization. This flaw could allow attackers to execute arbitrary code on affected systems. The vulnerability stems from improper handling of serialized data, enabling potential remote code execution. Users of OpenEMR should update to version 8.3.0 or later to patch this security weakness.
2026-11-15 NEW 2026Laravel: task output injection in Concurrency process (deserialization) intermediate RCELibrary for mitigating task output injection in Laravel's Concurrency processes. This vulnerability, identified as GHSA-63hm-g96r-2m2q, allows remote code execution when user-controlled text in a concurrency task is improperly handled, leading to the forging of the framework's internal result message. Exploitation requires insecure coding practices involving user input and specific output patterns to stdout, such as hand-built JSON via string concatenation. Patches are available in versions 13.40.0 and 12.71.0.
2026-11-15 NEW 2026RBH Integra32 deserialization (critical) intermediate RCELibrary for RBH Access Technologies Integra32 Security Management System 5.1.11 R5.0, identifying a critical unauthenticated deserialization vulnerability in the deprecated .NET Remoting service. Exploitable via BinaryServerFormatterSinkProvider at TypeFilterLevel.Full on TCP port 2001 using AxcvcLt.dll, attackers can leverage gadget payloads like ysoserial.net's TypeConfuseDelegate for arbitrary file read/write, credential coercion, and remote code execution as NT AUTHORITY\SYSTEM.
2026-11-14 2026CakePHP TwigView unserialize filter allows PHP object deserialization (CVE-2026-59192) news RCEA vulnerability in CakePHP's TwigView component, CVE-2026-59192, allows for PHP object deserialization due to an unrestricted `unserialize` filter. This can be exploited for remote code execution. The issue has been patched in versions 1.3.3 and 2.1.2.
2026-11-14 2026Unsafe Java deserialization in Nitrite legacy database migration (CVE-2026-64678) news 1 min read RCELibrary for addressing CVE-2026-64678, an unsafe Java deserialization vulnerability in Nitrite versions prior to 4.4.1. The flaw exists within the automatic migration of legacy v1 database stores, where unfiltered `ObjectInputStream.readObject()` can lead to arbitrary code execution if attacker-controlled serialized data and suitable gadget chains are present on the application's classpath. The library offers guidance on patching, interim mitigations like restricting database import, and an assume-compromise checklist.
2026-11-14 2026ZDI-26-951: FAIRChem deserialization of untrusted data RCE news Python RCEAdvisory ZDI-26-951 details a deserialization of untrusted data remote code execution vulnerability in Meta FAIRChem. This flaw allows remote attackers to execute arbitrary code by tricking users into opening a malicious checkpoint file, leading to deserialization of unsafe data. The vulnerability requires user interaction, as the target must visit a malicious page or open a malicious file to trigger the exploit within the affected installations of Meta FAIRChem. → zerodayinitiative.com
2026-11-14 2026ZDI-26-908: Meta Detectron2 deserialization of untrusted data RCE news Python RCEAdvisory for ZDI-26-908 details a Meta Detectron2 deserialization of untrusted data remote code execution vulnerability. Attackers can exploit this flaw, which lies in the parsing of pickle files lacking proper validation, to execute arbitrary code within the affected process. User interaction is required, typically by visiting a malicious page or opening a malicious file. The only mitigation strategy is to restrict interaction with the product. → zerodayinitiative.com
2026-11-14 2026ZDI-26-940: fairseq deserialization of untrusted data RCE news 1 min read AI Python RCEWriteup on ZDI-26-940, a deserialization of untrusted data RCE vulnerability in Meta fairseq's `dump_km_label` module. This flaw, affecting the `ApplyKmeans` class, allows remote code execution by tricking a user into visiting a malicious page or opening a crafted file. The fairseq repository is no longer maintained, making unrestricted interaction the only mitigation. → zerodayinitiative.com
2026-11-13 2026MongoDB C# Driver BSON type allow-list bypass (deserialization) newsLibrary for mitigating a BSON type allow-list bypass in the MongoDB C# Driver. This vulnerability, identified as GHSA-vjj2-chwg-8r8c, allows attackers to circumvent type-restriction checks during polymorphic data deserialization into loosely typed members. Exploitation can lead to arbitrary code execution by instantiating .NET types outside the intended allow-list, compromising confidentiality, integrity, and availability without requiring special privileges.
2026-11-13 2026Lenovo System Update deserialization newsA deserialization vulnerability in Lenovo System Update allows remote attackers to execute arbitrary code on affected systems. This flaw arises from improper handling of untrusted data during the deserialization process. Successful exploitation could lead to a complete system compromise. Users are advised to update their Lenovo System Update software to the latest version to mitigate this security risk.
2026-11-13 2026Apache XML Graphics FOP unsafe deserialization in hyphenation loading newsApache XML Graphics FOP is vulnerable to unsafe deserialization when loading hyphenation files. An attacker could exploit this by providing a malicious hyphenation file, leading to remote code execution on the affected server. This vulnerability arises from the improper handling of serialized data during the hyphenation loading process. Users are advised to update to a patched version of Apache XML Graphics FOP to mitigate this security risk.
2026-11-13 2026Telerik UI for WPF deserialization newsTelerik UI for WPF versions prior to 2026.4.1110 contain a deserialization vulnerability. This flaw allows an attacker to execute arbitrary code by crafting malicious input. Users are strongly advised to update to version 2026.4.1110 or later to patch this security risk.
2026-11-13 2026Telerik Reporting deserialization of untrusted data news RCETelerik Reporting versions prior to 20.3.26 are vulnerable to deserialization of untrusted data. This means that if an attacker can control the data being deserialized by the application, they could potentially execute arbitrary code. This vulnerability arises from insufficient validation of the data before it is processed by the Telerik Reporting components. Users are advised to update to version 20.3.26 or later to patch this security flaw.
2026-11-12 2026Apache Roller deserialization of untrusted data newsApache Roller is vulnerable to deserialization of untrusted data, allowing remote attackers to execute arbitrary code. This vulnerability arises from insufficient validation of serialized data. Successful exploitation could lead to full system compromise.
2026-11-12 2026Apache TomEE CVE-2026-111241: code execution via deserialized passivation files news RCEApache TomEE has a critical vulnerability, CVE-2026-111241, that allows for code execution. This exploit is possible through the deserialization of specially crafted passivation files. Attackers can leverage this flaw to compromise systems running vulnerable versions of Apache TomEE. Users are strongly advised to update to a patched version as soon as possible to mitigate this risk.
2026-11-12 2026Pickle-Me-Not: CTF pickle deserialization writeup intermediate 2 min read PythonWriteup of a CTF challenge involving static analysis of a Python pickle file. The solve requires reconstructing a SHA-256 keystream from diagnostic logs and a specific tensor view's offset, shape, and stride to decrypt a flag XORed with ciphertext. The challenge highlights a path-policy bypass in the gateway artifact endpoint and demonstrates the security implications of deserializing untrusted pickle data by emphasizing a "don't run it" approach.
2026-11-11 2026Visual Studio RCE via deserialization (CVE-2026-114151) news 3 min read RCEAdvisory for CVE-2026-114151 details a .NET vulnerability in `Microsoft.DiaSymReader.Native` versions earlier than 18.9.1-beta1.26515.1. A crafted PDB LF_DECIMAL payload can trigger an out-of-bounds stack write when processed via `ISymUnmanagedConstant.GetValue`, leading to high severity remote code execution on Windows platforms (x86, x64, Arm64) when attacker-supplied PDBs are processed. Affected components include .NET runtime packs and Windows SDK distributions.
2026-11-11 2026OpenVox CVE-2026-115731: Puppet report deserialization to server RCE news 6 min read RCEAnalysis of GHSA-ffr6-rwpm-p5wq and CVE-2026-115731 reveals a critical RCE vulnerability in OpenVox and Puppet. An authenticated attacker with a client certificate can exploit report deserialization, specifically the `Puppet::Pops::Serialization::FromDataConverter`, by crafting malicious `__ptype` fields within report events. This allows arbitrary Ruby code execution via `class_eval`, impacting critical infrastructure like banks and government datacenters. The vulnerability is exacerbated by common autosign configurations, effectively making it unauthenticated. The exploit grants code execution as the `puppet` user, enabling privilege escalation via access to the CA signing key.
2026-11-11 2026Windows DNS deserialization of untrusted data RCE intermediate RCEWindows DNS suffers from a deserialization vulnerability when processing untrusted data, enabling Remote Code Execution (RCE). This flaw allows attackers to potentially gain control of affected systems by crafting malicious data that the DNS server processes. Exploiting this vulnerability could lead to significant security breaches, compromising the integrity and confidentiality of Windows servers.
2026-11-11 2026Intel BigDL before 2.5.0b1 deserialization of untrusted data intermediate AI PythonLibrary for Intel BigDL, versions prior to 2.5.0b1, contains a deserialization of untrusted data vulnerability. An unprivileged adversary can exploit this flaw with low complexity and network access to escalate privileges, impacting confidentiality, integrity, and availability.
2026-11-10 2026rConfig deserialization (critical) news RCELibrary of techniques for exploiting rConfig Core 8.0.0 before 8.2.19 and rConfig Pro before 8.4.1 deserialization vulnerabilities. Attackers can achieve unauthenticated arbitrary code execution via crafted serialized jobs pushed to Redis-backed Laravel Horizon queues. Exploiting Laravel's CallQueuedHandler and the bundled laravel/tinker package allows for command execution, web shell deployment, and credential exfiltration. Authenticated users can also deliver payloads by targeting the rConfig server's loopback address if Redis is inaccessible externally.
2026-11-10 2026Exploiting a PHP object injection in Profile Builder Pro in the era of AI intermediate 10 min read AI RCEWriteup of CVE-2026-7647, a PHP object injection in Cozmoslabs' Profile Builder Pro, detailing how unauthenticated users can achieve remote code execution by exploiting insecure deserialization. The analysis highlights the process of identifying gadget chains, including a novel Monolog POP chain found in The-Event-Calendar plugin, enabling the deployment of a web shell. The technique involves crafting a malicious object within the `args` HTTP parameter of the `wppb_request_users_pins` AJAX action.
2026-11-10 2026ZDI-26-850: Microsoft Qlib MLflowRecorder deserialization RCE news 9 min read AI Python RCEReference detailing ZDI-26-850, a deserialization vulnerability in Microsoft Qlib's MLflowRecorder allowing RCE via user interaction. It outlines the exploit chain involving Python's `pickle` and provides detection logic for shell spawning and unexpected network connections from Python processes referencing `qlib` or `mlflow`, crucial for defending quantitative research and ML environments.

Frequently Asked Questions

What makes deserialization vulnerabilities so dangerous?
Deserialization vulnerabilities often lead directly to remote code execution (RCE) because the deserialization process can trigger arbitrary method calls through gadget chains — sequences of existing classes whose methods chain together to execute attacker-controlled commands. A single crafted payload can compromise an entire server.
What is a gadget chain?
A gadget chain is a sequence of existing classes and methods in an application's classpath that, when triggered during deserialization, produce a dangerous side effect like code execution. Tools like ysoserial (Java), phpggc (PHP), and peas (Python) generate payloads for known gadget chains in popular libraries and frameworks.
How do you prevent insecure deserialization?
Never deserialize untrusted data using native serialization formats (Java ObjectInputStream, Python pickle, PHP unserialize). Use safe data formats like JSON for data exchange. If native serialization is required, implement strict type allowlists, integrity checks (HMAC signatures), and consider using look-ahead deserialization that validates types before instantiation.

Weekly AppSec Digest

Get new resources delivered every Monday.