AI Security Threat Feed

Picklescan has a missing detection when calling built-in python profile.Profile.run

Picklescan has a missing detection when calling built-in python trace.Trace.runctx

Picklescan has a missing detection when calling built-in python trace.Trace.run

LlamaIndex affected by a Denial of Service (DOS) in JSONReader

XGrammar affected by Denial of Service by infinite recursion grammars

Picklescan missing detection when calling pytorch function torch.utils._config_module.load_config

Picklescan missing detection when calling pytorch function torch.jit.unsupported_tensor_ops.execWrapper

Picklescan missing detection when calling pytorch function torch.utils.data.datapipes.utils.decoder.basichandlers

Picklescan missing detection when calling pytorch function torch.utils.collect_env.run

Picklescan missing detection when calling pytorch function torch.fx.experimental.symbolic_shapes.ShapeEnv.evaluate_guards_expression

Picklescan missing detection when calling pytorch function torch._dynamo.guards.GuardBuilder.get

Picklescan missing detection when calling pytorch function torch.utils.bottleneck.__main__.run_cprofile

vLLM has remote code execution vulnerability in the tool call parser for Qwen3-Coder

Picklescan has pickle parsing logic flaw that leads to malicious pickle file bypass

ExecuTorch integer overflow vulnerability

ExecuTorch out-of-bounds access vulnerability

ExecuTorch vulnerable to Heap-based Buffer Overflow

ExecuTorch heap buffer overflow vulnerability

ExecuTorch integer overflow vulnerability

ExecuTorch vulnerable to Heap-based Buffer Overflow attack

LlamaIndex vulnerable to data loss through hash collisions in its DocugamiReader class

LlamaIndex vulnerable to Path Traversal attack through its encode_image function

LlamaIndex vulnerable to DoS attack through uncontrolled recursive JSON parsing

LlamaIndex has Incomplete Documentation of Program Execution related to JsonPickleSerializer component

gateway_proxy_handler in MLflow before 3.1.0 lacks gateway_path validation.

llama_index vulnerable to SQL Injection

vLLM vulnerable to Regular Expression Denial of Service

label-studio vulnerable to Cross-Site Scripting (Reflected) via the label_config parameter.

LlamaIndex Vulnerable to Denial of Service (DoS)

Browser Use allows bypassing `allowed_domains` by putting a decoy domain in http auth username portion of a URL

LLama Factory enables fine-tuning of large language models. Prior to version 1.0.0, a critical vulnerability exists in the `llamafy_baichuan2.py` script of the LLaMA-Factory project. The script...

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

vLLM vulnerable to Denial of Service by abusing xgrammar cache

xgrammar Vulnerable to Denial of Service (DoS) by abusing unbounded cache in memory

Picklescan missing detection when calling built-in python library function timeit.timeit()

Picklescan failed to detect to some unsafe global function in Numpy library

Picklescan Vulnerable to Exfiltration via DNS via linecache and ssl.get_server_certificate

SageMaker Workflow component allows possibility of MD5 hash collisions

LiteLLM Has an Improper Authorization Vulnerability

LiteLLM Reveals Portion of API Key via a Logging File

LiteLLM Vulnerable to Denial of Service (DoS) via Crafted HTTP Request

Open WebUI allows Remote Code Execution via Arbitrary File Upload to /audio/api/v1/transcriptions

Open Neural Network Exchange (ONNX) Path Traversal Vulnerability

PyTorch Lightning path traversal vulnerability

Open WebUI Unauthenticated Multipart Boundary Denial of Service (DoS) Vulnerability in api/chat/file

Open WebUI Cross-Site Request Forgery (CSRF) Vulnerability

LoLLMS Code Injection vulnerability

LlamaIndex Uncontrolled Resource Consumption vulnerability

A vulnerability in langchain-core versions >=0.1.17,<0.1.53, >=0.2.0,<0.2.43, and >=0.3.0,<0.3.15 allows unauthorized users to read arbitrary files from the host file system. The issue arises from...

A vulnerability in BerriAI/litellm, as of commit 26c03c9, allows unauthenticated users to cause a Denial of Service (DoS) by exploiting the use of ast.literal_eval to parse user input. This function...