Browse / Unsafe Model Loading and Supply Chain / Unsafe deserialisation
Unsafe deserialisation
77 records in Unsafe Model Loading and Supply Chain.
- vLLM: Public multimodal SHM handle forgery reaches worker-side `pickle.loads()` with insecure serialization disabled Forged shared-memory handles in vLLM's multimodal cache reach pickle.loads in the worker, and this works without the insecure-serialisation opt-in.
- NLTK: Allowlisted pickle loaders still permit code execution in current source Allowlisted pickle loaders trust whole module namespaces, so a crafted pickle can call dangerous helpers such as subprocess-spawning functions during unpickling.
- MLFLOW_ALLOW_PICKLE_DESERIALIZATION=False safety control bypassed by mlflow.statsmodels flavor — RCE via crafted model artifact The statsmodels loader skips the pickle deserialisation guard, so a crafted model artifact runs code when loaded even with pickle disabled.
- Feast vulnerable to arbitrary code execution via unsafe deserialization of dill-serialized UDFs Stored registry functions serialised with dill are deserialised by the server, so a malicious function placed in the registry runs code on the feature server.
- Keras: TorchModuleWrapper can deserialize unsafe PyTorch pickle data The TorchModuleWrapper config loader calls torch.load with unrestricted unpickling unless an ambient safe-mode flag is set, so untrusted configs run code.
- PyTorch Lightning allows arbitrary code execution through checkpoint _instantiator hyperparameters A crafted checkpoint supplies a module name in hyperparameters that is imported during loading, bypassing the weights_only protection.
- LangGraph Checkpoint: Unsafe JSON deserialization in checkpoint loading Tampered checkpoint bytes in a persistent store can make the JSON serializer rebuild arbitrary Python objects when a checkpoint is loaded.
- MLFlow Creates a Temporary File With Insecure Permissions Model artifact temp directories are created world- or group-writable, so local users can replace cloudpickle files that are later loaded and executed.
- SGLang: Unauthenticated RCE via --enable-custom-logit-processor A custom logit processor sent to the server is deserialised with dill.loads, so an unauthenticated request can run Python code when the option is enabled.
- LangSmith SDK: Public prompt pull deserializes untrusted manifests without trust boundary warning Pulling a public prompt by owner/name deserialises its manifest as trusted configuration, so attacker-authored constructor arguments and environment-backed secret references can run or leak.
- Ludwig framework is vulnerable to insecure deserialization through its predict() method. A crafted pickle file supplied as a dataset path is loaded with pandas.read_pickle without validation, so embedded Python objects run during prediction.
- Ludwig framework is vulnerable to insecure deserialization in its model serving component The model server loads weight files with unrestricted torch.load, so a crafted checkpoint executes arbitrary Python objects when served.
- PyTorch Lightning load_from_checkpoint has an insecure checkpoint deserialization load_from_checkpoint calls torch.load without weights_only, so a malicious checkpoint runs pickle-based code when it is loaded.
- LangChain vulnerable to unsafe deserialization of attacker-controlled objects through overly broad `load()` allowlists Overly permissive load settings revive attacker-supplied serialized constructor dictionaries in runtime paths, instantiating classes with attacker-chosen arguments.
- SGLang has an Improper Input Validation/Injection Issue A remotely supplied model or tokenizer reference causes unsafe deserialisation inside the transformers helper that loads the tokenizer.
- mem0ai mem0 has an Improper Input Validation Issue The FAISS vector store module calls pickle load on stored data, so a crafted file executes code when deserialised.
- HuggingFace Transformers allows for arbitrary code execution in the `Trainer` class A crafted RNG state checkpoint executes code when the Trainer loads it, because torch.load runs without the weights-only restriction.
- SGLang's encoder parallel disaggregation system is vulnerable to unauthenticated remote code execution through the disaggregation module Network-reachable disaggregation handlers pass untrusted bytes to pickle.loads without authentication, so a remote sender runs code on the server.
- LangGraph checkpoint loading has unsafe msgpack deserialization Persisted checkpoints are decoded with msgpack that can rebuild Python objects, so someone who can write to the store can get code run on load.
- PickleScan has multiple stdlib modules with direct RCE not in blocklist The scanner's blocklist omits several standard library modules that can run shell commands, so pickles using them are reported as clean.
- LangGraph: BaseCache Deserialization of Untrusted Data may lead to Remote Code Execution When msgpack fails, LangGraph's default cache serializer falls back to pickle.loads, so anyone able to write the cache backend can run code.
- Picklescan is vulnerable to RCE via missing detection when calling numpy.f2py.crackfortran.getlincoef A pickle that calls a NumPy F2PY helper to run Python code evades Picklescan's detection, so loading the scanned-as-safe file executes commands.
- Picklescan is vulnerable to RCE via missing detection when calling numpy.f2py.crackfortran.param_eval A pickle payload calls a numpy helper that the scanner does not flag, so scanning passes and loading the file executes attacker-supplied code.
- Picklescan missing detection when calling numpy.f2py.crackfortran.getlincoef Picklescan's blocklist omits numpy.f2py.crackfortran functions that eval strings, so a pickle calling getlincoef runs code while passing the scan.
- Hugging Face smolagents: Unsafe deserialization in Remote Python Executor leads to RCE The remote executor unpickles unvalidated network data, so an unauthenticated sender can run code as the service account.
- LangGraph Checkpoint affected by RCE in "json" mode of JsonPlusSerializer A fallback JSON serialisation mode lets crafted checkpoint payloads rebuild arbitrary objects during loading, running attacker-chosen Python code.
- Keras framework vulnerable to deserialization of untrusted data A crafted Keras file referencing a TorchModuleWrapper class runs attacker code while loading, even with safe mode enabled, whether the file is local or remote.
- Keras is vulnerable to Deserialization of Untrusted Data A crafted .keras archive re-enables unsafe deserialisation from inside its config before layer parsing, then a pickled Lambda layer runs attacker code when the model loads.
- The Keras `Model.load_model` method **silently** ignores `safe_mode=True` and allows arbitrary code execution when a `.h5`/`.hdf5` file is loaded. Loading an .h5 model ignores safe_mode=True, so pickled code inside the file runs on the victim's machine during load.
- Picklescan Bypass is Possible via File Extension Mismatch Picklescan sends a plain pickle with a PyTorch file extension down the PyTorch path, which errors out without running the pickle scanner, so malicious payloads go unflagged.
- Picklescan: ZIP archive scan bypass is possible through non-exhaustive Cyclic Redundancy Check Picklescan returns no results for a ZIP archive containing a bad CRC, while PyTorch can still load the hidden malicious pickle inside it.
- SGLang Remote Code Execution Vulnerability via Unsafe Deserialization in update_weights_from_tensor The weight-update endpoint deserialises attacker-supplied tensor payloads unsafely, letting a remote caller run code on the server.
- Picklescan is missing detection when calling built-in python idlelib.pyshell.ModifiedInterpreter.runcode A pickle payload calling idlelib's ModifiedInterpreter.runcode is not flagged by picklescan, so loading the malicious model file executes attacker code.
- Picklescan has a missing detection when calling built-in python idlelib.debugobj.ObjectTreeItem The pickle scanner does not flag calls to idlelib debugobj methods, so a malicious pickle that uses them passes the scan and runs commands when loaded.
- Picklescan has a missing detection when calling built-in python profile.Profile.runctx A pickle whose reduce calls profile.Profile.runctx runs arbitrary code when loaded, and the scanner does not flag that built-in function.
- Picklescan has a missing detection when calling built-in python idlelib.autocomplete.AutoComplete.get_entity A pickle that calls idlelib AutoComplete.get_entity executes attacker code on load, and the scanner misses this built-in callable.
- Picklescan has a missing detection when calling built-in python idlelib.autocomplete.AutoComplete.fetch_completions A pickle payload calling idlelib's AutoComplete.fetch_completions is not flagged by picklescan, so loading it runs attacker-chosen code.
- Picklescan has a missing detection when calling built-in python library idlelib.calltip.get_entity A pickle payload calls a standard library helper that evaluates strings as Python, and the scanner does not flag that call, so loading the file runs code.
- Picklescan has a missing detection when calling built-in python trace.Trace.runctx A pickle payload calling trace.Trace.runctx during reconstruction runs commands on load, and the scanner fails to flag it as dangerous.
- Picklescan is missing detection when calling built-in python idlelib.pyshell.ModifiedInterpreter.runcommand A pickle payload calls an idlelib function the scanner does not flag, so a malicious model passes scanning and runs code when loaded.
- Picklescan missing detection when calling pytorch function torch.utils.bottleneck.__main__.run_cprofile A pickle payload calls a torch profiling helper that runs arbitrary code, and the scanner fails to flag that call, so loading the file executes it.
- Keras vulnerable to CVE-2025-1550 bypass via reuse of internal functionality Crafted arguments to built-in Keras modules let a malicious saved model trigger file overwrites during loading even with safe mode enabled.
- Picklescan has pickle parsing logic flaw that leads to malicious pickle file bypass A malformed STACK_GLOBAL opcode sequence makes picklescan lose track of arguments, so malicious pickle files evade detection.
- SKOPS Card.get_model happily allows arbitrary code execution Card loading silently falls back from the safe skops format to joblib for non-zip files, so a crafted model file runs arbitrary code on load.
- Picklescan missing detection when calling built-in python library function timeit.timeit() A pickle payload calls timeit.timeit as a gadget to run shell commands, evading picklescan's denylist of dangerous globals.
- AgentScope Deserialization Vulnerability The RPC agent server calls dill.loads on client-supplied data during agent creation, so a crafted payload executes commands on the server.
- Kedro deserialization vulnerability A crafted shelve file contains a pickle payload that runs Python code when Kedro's ShelveStore reads the stored session data.
- vLLM Deserialization of Untrusted Data vulnerability The message queue dequeue path unpickles data received from sockets, so a crafted network message runs code on the receiving machine.
- vLLM Allows Remote Code Execution via Mooncake Integration vLLM's Mooncake KV-transfer pipe unpickles data received over a network socket, so a crafted payload sent to it runs code.
- Zip Flag Bit Exploit Crashes Picklescan But Not PyTorch Modifying ZIP flag bits makes picklescan skip a malicious pickle inside a PyTorch archive, while torch.load still reads and executes it.