Deserialization of Untrusted Data Affecting mlflow package, versions [2.1.0,3.15.0)

mlflow is a platform to streamline machine learning development, including tracking experiments, packaging code into reproducible runs, and sharing and deploying models.

Affected versions of this package are vulnerable to Deserialization of Untrusted Data in the _load_model() and _load_pyfunc() functions of the mlflow.statsmodels flavor (mlflow/statsmodels/__init__.py), which call statsmodels.iolib.api.load_pickle() without honoring the MLFLOW_ALLOW_PICKLE_DESERIALIZATION control that guards the other flavors. An attacker can achieve arbitrary code execution by placing a malicious pickle file alongside an MLmodel artifact that declares the statsmodels flavor in an accessible artifact store, which executes when a victim calls mlflow.pyfunc.load_model() on it. This requires the attacker to write the artifact into a store the victim loads from, and the victim to load that specific model.

Serialization is a process of converting an object into a sequence of bytes which can be persisted to a disk or database or can be sent through streams. The reverse process of creating object from sequence of bytes is called deserialization. Serialization is commonly used for communication (sharing objects between multiple hosts) and persistence (store the object state in a file or a database). It is an integral part of popular protocols like Remote Method Invocation (RMI), Java Management Extension (JMX), Java Messaging System (JMS), Action Message Format (AMF), Java Server Faces (JSF) ViewState, etc.

Deserialization of untrusted data (CWE-502) is when the application deserializes untrusted data without sufficiently verifying that the resulting data will be valid, thus allowing the attacker to control the state or the flow of the execution.

• GitHub Advisory
• GitHub Commit
• GitHub PR
• GitHub Release