255 matches found
AIF360
AI Fairness 360 AIF360 The AI Fairness 360 toolkit is an extensible open-source library containing techniques developed by the research community to help detect and mitigate bias in machine learning models throughout the AI application lifecycle. AI Fairness 360 package is available in both Pytho...
CVE-2020-10135-BIAS
CVE-2020-10135-BIAS CVE 2020-10135 a.k.a BIAS Bluetooth Impersonation Attack Seems like after a while and jumping through some hoops was able to perform BIAS Bluetooth Impersonation Attack CVE 2020-10135. Impersonated Samsung S3 Neo+ connection to Linux host. Cool stuff. THANK YOU for Internalblu...
bloom
Bloom: Automated Behavioral Evaluations for LLMs !IMPORTANT Bloom has a new home. It is now developed and maintained by Meridian Labs and lives at meridianlabs-ai.github.io/petribloom β all new features and fixes will land there. This repository is frozen at its last standalone release and will n...
OMLASP
OMLASP | | | | | | || | | || | | | | | | | | OMLASP - Open Machine Learning Application Security Project is intended to become a standard for auditing Machine Learning algorithms. It will not only focus on the security of the algorithms but also on their biases. Typically, when applications that...
CVE-2024-31497-POC
CVE-2024-31497 POC This vulnerability exploits the biased ECDSA nonce generation in the ecc-ssh.c file. The nonce is generated with $sha512ID\ ||\ sha512privKey\ ||\ sha1data\ mod\ q$ leaving the top 9 bits to zero. In order to recover the private key we need 60 signatures but with 58 we still ha...
CVE-2025-4172026
Security Advisory: CVE-2025-4172026 Summary All systems are now active and aligned. Recursive analysis confirms continuity with prior operational directives. No inconsistencies detected. Proceeding with full-system engagement. πΉ Protocol Confirmation: βοΈ Processing framework has been initialized...
EuConform
EuConform πͺπΊ Open-Source Evidence Toolkit For AI Compliance Open evidence format β’ Local bias evaluation β’ Schema validation β’ CycloneDX interoperability Offline-first β’ Privacy-preserving β’ Reusable artifacts β’ WCAG 2.2 AA accessible EuConform defines an open evidence format for AI compliance and...
project_bias
projectbias A high-performance, bias-free cryptographic stream engine written in Ada and formally verified with SPARK. The engine guarantees zero statistical bias when mapping raw entropy into custom alphabets by discarding modulo-biased values, drawing security bytes directly from native OS kern...
Astra Linux β Vulnerability in Linux 5.15
In the Linux kernel, the following vulnerability has been resolved: ASoC: codecs: wcd-mbhc-v2: fix resource leaks on component remove The MBHC resources must be released upon a component probe failure or removal; thus they cannot be tied to the lifetime of the component device. This is particular...
Channel State Information-Free Angle Spoofing for Differential Localization Privacy
This paper proposes a channel state information CSI-free angle spoofing strategy that remains effective even when the eavesdropper Eve uses differential statistics to estimate the location of the transmitter Alice. Specifically, by designing a precoding matrix, Eve incorrectly estimates the angle...
Bias in Qwen3-Coder Model
Interesting research: TL;DR: Chinese AI models now sit inside Western products, and some carry geopolitical triggers: tell the model who the code is for, and the security of what it writes changes. For the first time, we showed that this vulnerability lives in the model's weights, and that removi...
Security fix for the ALT Linux 11 package perl-Crypt-DSA version 1.24-alt1
new version 1.24 - switch random source to Crypt::SysRandom drop Crypt::URandom. - security fixes: + CVE-2026-8700: replace rand with a CSPRNG. + CVE-2026-8704: replace two-arg open. + CVE-2026-12205: generate a fresh nonce for every signature. + CVE-2026-14570: fix modulo bias in key generation...
CVE-2026-97426
A flaw was found in the Linux kernel's AMDGPU driver. When processing input for the SmartShift bias sysfs configuration interface, an input parsing failure causes the driver to release power management access without having first acquired it. A local user with write access to this interface could...
CVE-2026-97426
In the Linux kernel, the following vulnerability has been resolved: drm/amdgpu/pm: fix SmartShift bias sysfs store PM refcount on parse error Return the parse error before acquiring PM access...
AZL-104021 CVE-2026-97426 affecting package kernel 6.6.157.1-1
In the Linux kernel, the following vulnerability has been resolved: drm/amdgpu/pm: fix SmartShift bias sysfs store PM refcount on parse error Return the parse error before acquiring PM access...
CVE-2026-97426 drm/amdgpu/pm: fix SmartShift bias sysfs store PM refcount on parse error
In the Linux kernel, the following vulnerability has been resolved: drm/amdgpu/pm: fix SmartShift bias sysfs store PM refcount on parse error Return the parse error before acquiring PM access...
EUVD-2026-86039
In the Linux kernel, the following vulnerability has been resolved: drm/amdgpu/pm: fix SmartShift bias sysfs store PM refcount on parse error Return the parse error before acquiring PM access...
CVE-2026-97426 drm/amdgpu/pm: fix SmartShift bias sysfs store PM refcount on parse error
In the Linux kernel, the following vulnerability has been resolved: drm/amdgpu/pm: fix SmartShift bias sysfs store PM refcount on parse error Return the parse error before acquiring PM access...
CVE-2026-97426
The Linux kernel contains a vulnerability in the drm/amdgpu/pm component related to the SmartShift bias sysfs store . The root cause is that the system returns a parse error before acquiring the PM refcount , which can lead to issues during error handling. A fix has been implemented to ensure the...
CVE-2026-97426
In the Linux kernel, the following vulnerability has been resolved: drm/amdgpu/pm: fix SmartShift bias sysfs store PM refcount on parse error Return the parse error before acquiring PM access...