CVE-2020-25502

A vulnerability in the installer component of Cisco AnyConnect Secure Mobility Client for Windows could allow an authenticated local attacker to copy user-supplied files to system level directories with system level privileges. The vulnerability is due to the incorrect handling of directory paths. An attacker could exploit this vulnerability by creating a malicious file and copying the file to a system directory. An exploit could allow the attacker to copy malicious files to arbitrary locations with system level privileges. This could include DLL pre-loading, DLL hijacking, and other related attacks. To exploit this vulnerability, the attacker needs valid credentials on the Windows system.

A vulnerability in the interprocess communication (IPC) channel of Cisco AnyConnect Secure Mobility Client for Windows could allow an authenticated, local attacker to perform a DLL hijacking attack. To exploit this vulnerability, the attacker would need to have valid credentials on the Windows system. The vulnerability is due to insufficient validation of resources that are loaded by the application at run time. An attacker could exploit this vulnerability by sending a crafted IPC message to the AnyConnect process. A successful exploit could allow the attacker to execute arbitrary code on the affected machine with SYSTEM privileges. To exploit this vulnerability, the attacker would need to have valid credentials on the Windows system.

An issue in Plone Docker Official Image 5.2.13 (5221) open-source software that could allow for remote code execution due to a package listed in ++plone++static/components not existing in the public package index (npm).

Outbyte PC Repair Installation File 1.7.112.7856 is vulnerable to Dll Hijacking. iertutil.dll is missing so an attacker can use a malicious dll with same name and can get admin privileges.

CyberGhostVPN Windows Client before v8.3.10.10015 was discovered to contain a DLL injection vulnerability via the component Dashboard.exe.

vLLM is an inference and serving engine for large language models (LLMs). Prior to 0.22.1, the vLLM Dockerfile is vulnerable to a dependency confusion attack through the flashinfer-jit-cache package. The package is installed from a custom index (flashinfer.ai/whl/) using --extra-index-url, but the package name was not registered on PyPI, and UV_INDEX_STRATEGY="unsafe-best-match" is set globally. An attacker who registers flashinfer-jit-cache on PyPI with version 0.6.11.post2 can execute arbitrary code as root during the Docker build and backdoor every resulting container image, enabling exfiltration of all user prompts, API credentials, and model data from production vLLM deployments This vulnerability is fixed in 0.22.1.