pickle
Safeguard articles tagged "pickle" — guides, analysis, and best practices for software supply chain and application security.
7 articles
Recurring vulnerability patterns in the PyPI ecosystem
PyYAML shipped two rounds of deserialization fixes in under two years — CVE-2017-18342 and CVE-2020-14343 — because the underlying pattern kept resurfacing.
The Python pickle security model, explained
Python's own docs warn that unpickling can execute arbitrary code — yet pickle is still the default weight format behind millions of ML model downloads.
Pickle and Deserialization Security in Python
Unpickling untrusted data is arbitrary code execution, by design. Here is why pickle is dangerous, where it hides, and what to use instead.
Hugging Face Model Hub Supply Chain Risks in 2025
Pickle deserialization, malicious Spaces, and namespace squatting: what 2024-2025 taught us about the Hugging Face model supply chain.
Hugging Face's Guardian-Plus-Picklescan Stack: How the Model Hub Scanning Posture Evolved Through 2025-2026
Following NullifAI and the broken-pickle bypass campaigns, Hugging Face layered Protect AI's Guardian on top of Picklescan, ClamAV, and secrets scanning across 1.5 million public models. Here is the defender view of the new pipeline.
Hugging Face Pickle Backdoor Research 2025
Pickle-serialized model files remain a live attack surface on Hugging Face. Here is what 2025 research disclosed about persistent backdoors and what defenders should do about it.
nullifAI: Broken Pickles and the Hugging Face Detection Gap
ReversingLabs disclosed two malicious Hugging Face models that evaded Picklescan by using broken 7z-packed PyTorch archives. We unpack the technique.