Deserialization of Untrusted Data

The Keras Model.load_model method can be exploited to achieve arbitrary code execution, even with safe_mode=True. One can create a specially crafted .keras model archive that, when loaded via Model.load_model, will trigger arbitrary code to be executed. This is achieved by crafting a special config.json (a file within the .keras archive) that will invoke keras.config.enable_unsafe_deserialization() to disable safe mode. Once safe mode is disable, one can use the Lambda layer feature of keras, which allows arbitrary Python code in the form of pickled code. Both can appear in the same archive. Simply the keras.config.enable_unsafe_deserialization() needs to appear first in the archive and the Lambda with arbitrary code needs to be second.Show less

A deserialization vulnerability in the License Servlet of Fortra's GoAnywhere MFT allows an actor with a validly forged license response signature to deserialize an arbitrary actor-controlled object, possibly leading to command injection.Show less

Greenshot is an open source Windows screenshot utility. Greenshot 1.3.300 and earlier deserializes attacker-controlled data received in a WM_COPYDATA message using BinaryFormatter.Deserialize without prior validation or authentication, allowing a local process at the same integrity level to trigger arbitrary code execution inside the Greenshot process. The vulnerable logic resides in a WinForms WndProc handler for WM_COPYDATA (message 74) that copies the supplied bytes into a MemoryStream and invokes BinaryFormatter.Deserialize, and only afterward checks whether the specified channel is authorized. Because the authorization check occurs after deserialization, any gadget chain embedded in the serialized payload executes regardless of channel membership. A local attacker who can send WM_COPYDATA to the Greenshot main window can achieve in-process code execution, which may aid evasion of application control policies by running payloads within the trusted, signed Greenshot.exe process. This issue is fixed in version 1.3.301. No known workarounds exist.Show less

A Java deserialisation vulnerability has been discovered in Jaspersoft Library. Improper handling of externally supplied data may allow attackers to execute arbitrary code remotely on systems that use the affected libraryShow less

A vulnerability in Apache Fory allows a remote attacker to cause a Denial of Service (DoS). The issue stems from the insecure deserialization of untrusted data. An attacker can supply a large, specially crafted data payload that, when processed, consumes an excessive amount of CPU resources during the deserialization process. This leads to CPU exhaustion, rendering the application or system using the Apache Fory library unresponsive and unavailable to legitimate users. Users of Apache Fory are strongly advised to upgrade to version 0.12.2 or later to mitigate this vulnerability. Developers of libraries and applications that depend on Apache Fory should update their dependency requirements to Apache Fory 0.12.2 or later and release new versions of their software.Show less

Dataease is an open source data analytics and visualization platform. In Dataease versions up to 2.10.12 the H2 data source implementation (H2.java) does not verify that a provided JDBC URL starts with jdbc:h2. This lack of validation allows a crafted JDBC configuration that substitutes the Amazon Redshift driver and leverages the socketFactory and socketFactoryArg parameters to invoke org.springframework.context.support.FileSystemXmlApplicationContext or ClassPathXmlApplicationContext with an attacker‑controlled remote XML resource, resulting in remote code execution. Versions up to and including 2.10.12 are affected. The issue is fixed in version 2.10.13. Updating to version 2.10.13 or later is the recommended remediation. No known workarounds exist.Show less

Dataease is an open-source data visualization and analysis platform. In versions up to and including 2.10.12, the Impala data source is vulnerable to remote code execution due to insufficient filtering in the getJdbc method of the io.dataease.datasource.type.Impala class. Attackers can construct malicious JDBC connection strings that exploit JNDI injection and trigger RMI deserialization, ultimately enabling remote command execution. The vulnerability can be exploited by editing the data source and providing a crafted JDBC connection string that references a remote configuration file, leading to RMI-based deserialization attacks. This issue has been patched in version 2.10.13. It is recommended to upgrade to the latest version. No known workarounds exist for affected versions.Show less

A vulnerability was determined in 1Panel-dev MaxKB up to 2.0.2/2.1.0. This issue affects some unknown processing of the file /admin/api/workspace/default/tool/debug. Executing manipulation of the argument code can lead to deserialization. The attack can be executed remotely. The exploit has been publicly disclosed and may be utilized. Upgrading to version 2.1.1 is capable of addressing this issue. It is suggested to upgrade the affected component.Show less

A flaw has been found in SEAT Queue Ticket Kiosk up to 20250827. This affects an unknown part of the component Java RMI Registry Handler. This manipulation causes deserialization. The attack can only be done within the local network. The attack is considered to have high complexity. It is indicated that the exploitability is difficult. The vendor was contacted early about this disclosure but did not respond in any way.Show less

A security flaw has been discovered in lmsys sglang 0.4.6. Affected by this vulnerability is the function main of the file /update_weights_from_tensor. The manipulation of the argument serialized_named_tensors results in deserialization. The attack can be launched remotely. The exploit has been released to the public and may be exploited. The vendor was contacted early about this disclosure but did not respond in any way.Show less

An unauthenticated attacker can trick a local user into executing arbitrary commands by opening a deliberately manipulated project file with an affected engineering tool. These arbitrary commands are executed in the user context.Show less

Due to a deserialization vulnerability in SAP NetWeaver, an unauthenticated attacker could exploit the system through the RMI-P4 module by submitting malicious payload to an open port. The deserialization of such untrusted Java objects could lead to arbitrary OS command execution, posing a high impact to the application's confidentiality, integrity, and availability.Show less

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.Show less