CVE-2025-66222

DeepChat is a smart assistant uses artificial intelligence. In 0.5.0 and earlier, there is a Stored Cross-Site Scripting XSS vulnerability in the Mermaid diagram renderer allows an attacker to execute arbitrary JavaScript within the application context. By leveraging the exposed Electron IPC...

Remote Command Execution

scio-pypi is vulnerable to Remote Command Execution. The vulnerability is due to torch.load executing unsafe deserialization even when weightsonly=True, which allows an attacker to craft malicious model files that trigger arbitrary code execution during loading...

CVE-2025-64110 Cursor: Authentication Bypass Possible via New Cursorignore Write

Cursor is a code editor built for programming with AI. In versions 1.7.23 and below, a logic bug allows a malicious agent to read sensitive files that should be protected via cursorignore. An attacker who has already achieved prompt injection, or a malicious model, could create a new cursorignore...

PT-2025-45062

Name of the Vulnerable Software and Affected Versions Cursor versions 1.7.44 and below Description Cursor, a code editor for programming with AI, has an issue where NTFS path quirks can be exploited by an attacker to bypass file protections and overwrite files that normally require user...

EUVD-2025-7036

Malicious code in bioql PyPI...

EUVD-2025-25170

Malicious code in bioql PyPI...

Keras safe_mode bypass allows arbitrary code execution when loading a malicious model.

...

Arbitrary Code Execution (ACE)

skops is vulnerable to Arbitrary Code Execution ACE. The vulnerability is due to Card.getmodel falling back to joblib for non-.zip file formats without warning, which allows an attacker to load a malicious model file and execute arbitrary code...

CVE-2025-50461

A deserialization vulnerability exists in Volcengine's verl 3.0.0, specifically in the scripts/modelmerger.py script when using the "fsdp" backend. The script calls torch.load with weightsonly=False on user-supplied .pt files, allowing attackers to execute arbitrary code if a maliciously crafted...

CVE-2025-50461

A deserialization vulnerability exists in Volcengine's verl 3.0.0, specifically in the scripts/modelmerger.py script when using the "fsdp" backend. The script calls torch.load with weightsonly=False on user-supplied .pt files, allowing attackers to execute arbitrary code if a maliciously crafted...

Exploit for CVE-2024-3660

CVE-2024-3660 – TensorFlow Keras Arbitrary Code Execution via...

CVE-2025-49847 llama.cpp Vulnerable to Buffer Overflow via Malicious GGUF Model

llama.cpp is an inference of several LLM models in C/C++. Prior to version b5662, an attacker‐supplied GGUF model vocabulary can trigger a buffer overflow in llama.cpp’s vocabulary‐loading code. Specifically, the helper trycopy in llama.cpp/src/vocab.cpp: llamavocab::impl::tokentopiece casts a ve...

Mind the Gap: a Practical Attack on GGUF Quantization

With the increasing size of frontier LLMs, post-training quantization has become the standard for memory-efficient deployment. Recent work has shown that basic rounding-based quantization schemes pose security risks, as they can be exploited to inject malicious behaviors into quantized models tha...

Machine Learning Models Have a Supply Chain Problem

Powerful machine learning ML models are now readily available online, which creates exciting possibilities for users who lack the deep technical expertise or substantial computing resources needed to develop them. On the other hand, this type of open ecosystem comes with many risks. In this paper...

CVE-2025-24357 Malicious model remote code execution fix bypass with PyTorch < 2.6.0

Description https://github.com/vllm-project/vllm/security/advisories/GHSA-rh4j-5rhw-hr54 reported a vulnerability where loading a malicious model could result in code execution on the vllm host. The fix applied to specify weightsonly=True to calls to torch.load did not solve the problem prior to...

NULL Pointer Dereference

Overview Affected versions of this package are vulnerable to NULL Pointer Dereference in gguf.go, which allows attackers to crash the application by supplying a malicious GGUF model as input. Note: This vulnerability has also been published as CVE-2025-0312. Details Denial of Service DoS describe...

CVE-2024-12055 DoS using malicious gguf model file in ollama/ollama

A vulnerability in Ollama versions =0.3.14 allows a malicious user to create a customized gguf model file that can be uploaded to the public Ollama server. When the server processes this malicious model, it crashes, leading to a Denial of Service DoS attack. The root cause of the issue is an...

CVE-2025-1433

A maliciously crafted MODEL file, when parsed through Autodesk AutoCAD, can force an Out-of-Bounds Read vulnerability. A malicious actor can leverage this vulnerability to cause a crash, read sensitive data, or execute arbitrary code in the context of the current process...

Insufficient Verification Of Data Authenticity

PickleScan is vulnerable to Insufficient Verification of Data Authenticity. The vulnerability is due to a discrepancy in filename handling due to differences between ZIP header filenames and directory listing filenames, which allows an attacker to bypass detection by causing PickleScan to crash...

Autodesk AutoCAD 缓冲区错误漏洞

Autodesk AutoCAD is a suite of professional 3D drawing software from Autodesk, Inc. A security vulnerability exists in Autodesk AutoCAD that originates from parsing a malicious specially crafted MODEL file that could result in an out-of-bounds read, which could cause a crash, read sensitive data,...