EUVD-2026-83800
GHSA-rh54-3493-vqgc
Severity by source
CVSS:4.0/AV:N/AC:L/AT:N/PR:N/UI:P/VC:L/VI:L/VA:L/SC:N/SI:N/SA:N/E:P/CR:X/IR:X/AR:X/MAV:X/MAC:X/MAT:X/MPR:X/MUI:X/MVC:X/MVI:X/MVA:X/MSC:X/MSI:X/MSA:X/S:X/AU:X/R:X/V:X/RE:X/U:X
Pickle deserialization yields arbitrary code execution (C/I/A:H), delivered over the network but requiring the victim to load a crafted file (UI:R); no privileges needed.
Primary rating from Vendor (VulDB).
CVSS VectorVendor: VulDB
Lifecycle Timeline
4DescriptionCVE.org
A security vulnerability has been detected in DLR-RM stable-baselines3 up to 2.9.0. This affects the function PPO.load/load_replay_buffer/VecNormalize.load of the file save_util.py. Such manipulation leads to deserialization. It is possible to launch the attack remotely. The exploit has been disclosed publicly and may be used. In v2.9.0 the PyTorch tensor load path is hardened (weights_only=True), but that hardening was later reverted on master via PR #1913 "Hotfix: revert loading with weights_only=True" [blocked] to fix PyTorch 1.13 compat - so even the one "safe" path is inconsistent across versions. #2281 was closed as a duplicate of #1831 since both are unsafe pickle deserialization - but #1831's fix (PR #41) only gated the Hugging Face Hub loader in the separate huggingface_sb3 package. This finding covers the core stable_baselines3 load APIs (PPO.load, load_replay_buffer, VecNormalize.load), which have no safe mode or gate and remained exploitable in v2.9.0 until the outstanding hardening (PR #2264) ships.
AnalysisAI
Unsafe pickle deserialization in the core load APIs of DLR-RM stable-baselines3 up to 2.9.0 lets an attacker who supplies a crafted model, replay buffer, or VecNormalize artifact achieve arbitrary code execution when a victim loads it via PPO.load, load_replay_buffer, or VecNormalize.load. The flaw is a CWE-502 pickle deserialization issue with no safe-mode gate or opt-in guard in these functions, and publicly available exploit code exists; the practical attack path is sharing or publishing a poisoned reinforcement-learning checkpoint that the victim is induced to load. …
Unlock full vulnerability intelligence
- Risk assessment & exploitation conditions
- Attack chain visualization
- Remediation with exact patch versions
- Threat intelligence from 22 sources
- Personal watchlist & email alerts
No credit card · 7-day full trial
Attack ChainAIDerived
Hypothetical attack flow derived from CVE metadata
Vulnerability AssessmentAI
| Exploitation | Exploitation requires a victim to load an attacker-controlled file through one of the core stable-baselines3 APIs - PPO.load (or any algorithm's load), load_replay_buffer, or VecNormalize.load - which internally deserialize via pickle in save_util.py with no safe-mode gate up through v2.9.0. … Additional conditions and limiting factors are described in the full assessment. |
| Risk Assessment | This is a classic Python pickle deserialization flaw (CWE-502) in the core load APIs of stable-baselines3 - PPO.load, load_replay_buffer, and VecNormalize.load in save_util.py. … Full risk analysis with EPSS, KEV, and SSVC signal comparison available after sign-in. |
| Exploit Scenario | Full exploit scenario with step-by-step reproduction available after sign-in. |
| Remediation | No vendor-released patch is identified at time of analysis for the core load APIs: the v2.9.0 weights_only=True hardening covered only the PyTorch tensor path and was reverted on master by PR #1913, while the outstanding hardening tracked as PR #2264 has not shipped - so do not rely on 2.9.0 behavior as safe. … Detailed patch versions, workarounds, and compensating controls in full report. |
Threat intelligence, references, and detailed analysis are available after sign-in.
Wazuh SIEM platform versions 4.4.0 through 4.9.0 contain an unsafe deserialization vulnerability in the DistributedAPI t
BentoML version 1.4.2 and earlier contains an unauthenticated remote code execution vulnerability through insecure deser
pgAdmin 4 contains critical remote code execution vulnerabilities in the Query Tool download and Cloud Deployment endpoi
The renderLocalView function in render/views.py in graphite-web in Graphite 0.9.5 through 0.9.10 uses the pickle Python
BentoML is a Python library for building online serving systems optimized for AI apps and model inference. Rated critica
OpenSSL before 0.9.8za, 1.0.0 before 1.0.0m, and 1.0.1 before 1.0.1h does not properly restrict processing of ChangeCiph
pyLoad download manager version prior to 0.5.0b3.dev77 exposes the Flask SECRET_KEY through an unauthenticated endpoint.
Langflow (a visual LLM pipeline builder) contains a critical unauthenticated code execution vulnerability (CVE-2026-3301
In Mercurial before 4.1.3, "hg serve --stdio" allows remote authenticated users to launch the Python debugger, and conse
Unauthenticated remote code execution affects Kestra OSS (the open-source event-driven orchestration platform) prior to
Unauthenticated remote code execution in Marimo ≤0.20.4 allows attackers to execute arbitrary system commands via the `/
pyLoad is the free and open-source Download Manager written in pure Python. Rated medium severity (CVSS 5.3), this vulne
Same weakness CWE-502 – Deserialization of Untrusted Data
View allSame technique Deserialization
View allShare
External POC / Exploit Code
Leaving vuln.today