Database/AI/ML frameworks & serving
ClearML client SDK: Deserialization of untrusted data
Impact
Deserialization of untrusted data — a malicious uploaded artifact executes code on the consuming host
Who can reach it
Customer-supplied artifact pulled by another user's ClearML job
What to do
Upgrade past 1.14.2. Cross-tenant if a shared ClearML server serves multiple teams
Fleet impact
How widespread
Common - ClearML is a widely used MLOps/experiment platform beside GPU clusters; affects SDK 0.17.0-1.14.2
Cost to remediate
**Image rebuild** of every image containing the ClearML SDK, plus revocation of every artifact in the store
Why it hits the whole fleet
Artifact.get() pickle-deserializes a maliciously uploaded artifact, so one poisoned artifact runs code on every user or job that reads it - a supply-chain fan-out inside the platform, with 10+ public PoCs
References
Related entries
- ClearML client SDK: Path traversal — a malicious dataset writes arbitrary files on the consumerCVE-2024-24591 · ClearML client SDKHigh
- Keras (`utils.get_file`): Path traversal in tar extraction in 3.11.3 (incomplete fix)CVE-2025-12638 · Keras (`utils.get_file`)High
- vLLM (multi-node ZeroMQ): Secondary vLLM host trusts unauthenticated ZeroMQ messagesCVE-2025-30165 · vLLM (multi-node ZeroMQ)High
- Kubeflow Community Distribution: Insecure default in the platform installCVE-2026-47237 · Kubeflow Community DistributionHigh
- PyTorch (mobile interpreter): Use-after-free in `torch/csrc/jit/mobile/interpreter.cpp`CVE-2024-31583 · PyTorch (mobile interpreter)High
- langchain-experimental (Python REPL): Python REPL exposed without an opt-inCVE-2024-38459 · langchain-experimental (Python REPL)High
This entry is curated: imported from vendor advisories with machine assistance, not yet individually verified. Confirm against your vendor's advisory before acting, and report anything wrong.