Undetectable Backdoors in Machine-Learning Models

New paper: "Planting Undetectable Backdoors in Machine Learning Models": Abstract: Given the computational cost and technical expertise required to train machine learning models, users may delegate the task of learning to a service provider. We show how a malicious learner can plant an undetectab...

Microsoft Obtains Court Order to Take Down Domains Used to Target Ukraine

Microsoft on Thursday disclosed that it obtained a court order to take control of seven domains used by APT28, a state-sponsored group operated by Russia's military intelligence service, with the goal of neutralizing its attacks on Ukraine. "We have since re-directed these domains to a sinkhole...

adversarial-robustness-toolbox (1.2.0), aristote-malo (1.9.5) +79 more potentially affected by CVE-2022-22816 via pillow (7.0.0)

pillow PYPI version 7.0.0 is affected by a known vulnerability. The following packages have a transitive dependency on pillow and may be impacted: - adversarial-robustness-toolbox 1.2.0 - aristote-malo 1.9.5 - auto-annotate =1.0.0, =1.0.2, =0.1.0.dev31, =1.0.1, =0.2.3, =0.0.1.25, =0.2.1, =0.0.2,...

Iran's Lyceum Hackers Target Telecoms, ISPs in Israel, Saudi Arabia, and Africa

A state-sponsored threat actor allegedly affiliated with Iran has been linked to a series of targeted attacks aimed at internet service providers ISPs and telecommunication operators in Israel, Morocco, Tunisia, and Saudi Arabia, as well as a ministry of foreign affairs MFA in Africa, new finding...

The features all Incident Response Plans need to have

By Paul Lee, Yuri Kramarz and Martin Lee. Adversaries are always growing their capabilities and changing their tactics, leading to a greater number of incidents and data breaches. This is supported by organizations such as ITRC who reports that the number of data breaches in 2021 is already...

CVE-2021-42694

An issue was discovered in the character definitions of the Unicode Specification through 14.0. The specification allows an adversary to produce source code identifiers such as function names using homoglyphs that render visually identical to a target identifier. Adversaries can leverage this to...

CVE-2021-42694

An issue was discovered in the character definitions of the Unicode Specification through 14.0. The specification allows an adversary to produce source code identifiers such as function names using homoglyphs that render visually identical to a target identifier. Adversaries can leverage this to...

Security feature bypass

DISPUTED An issue was discovered in the character definitions of the Unicode Specification through 14.0. The specification allows an adversary to produce source code identifiers such as function names using homoglyphs that render visually identical to a target identifier. Adversaries can leverage...

Recovering Real Faces from Face-Generation ML System

New paper: "This Person Probably Exists. Identity Membership Attacks Against GAN Generated Faces. Abstract: Recently, generative adversarial networks GANs have achieved stunning realism, fooling even human observers. Indeed, the popular tongue-in-cheek website http://thispersondoesnotexist.com,...

Attracting flies with Honey(gain): Adversarial abuse of proxyware

By Edmund Brumaghin and Vitor Ventura. With internet-sharing applications, or "proxyware," users download software that allows them to share a percentage of their bandwidth with other internet users for a fee, with the companies that created this software acting as a go-between.As proxyware has...

accuinsight (>=1.0.47, <=3.0.0), adapt-diagnostics (>=1.2.0, <=1.6.0) +332 more potentially affected by CVE-2021-37676 via tensorflow (>=2.0.0, <=2.3.2)

tensorflow PYPI version =2.0.0, =1.0.47, =1.2.0, =0.0.1, =0.5.6, =0.5.0, =0.1.0, =0.10.0.dev0, =2.0.0b0, =0.5.0, =1.1.0 and more Source cves: CVE-2021-37676 Source advisory: OSV:GHSA-V768-W7M9-2VMM...

‘Optical Adversarial Attack’ uses low-cost projector to trick AI

By Sudais Asif In the latest, we have another piece of research that deals with strikingly similar details but incorporating the trickery of Artificial Intelligence AI. This is a post from HackRead.com Read the original post: Optical Adversarial Attack uses low-cost projector to trick AI...

accuinsight (>=1.0.62, <=3.0.0), adapt-diagnostics (>=1.2.0, <=1.6.0) +95 more potentially affected by CVE-2021-37668 via tensorflow (>=2.3.0, <=2.3.2)

tensorflow PYPI version =2.3.0, =1.0.62, =1.2.0, =0.0.1a0, =0.0.1, =1.0.0rc1, =20210206.0.0, =0.1.0.dev1, =0.2.4, =1.0.1.0, =0.1.0, =2.0.0rc1, =2.0.3 and more Source cves: CVE-2021-37668 Source advisory: OSV:PYSEC-2021-290...

accuinsight (>=1.0.62, <=3.0.0), adapt-diagnostics (>=1.2.0, <=1.6.0) +95 more potentially affected by CVE-2021-37646 via tensorflow (>=2.3.0, <=2.3.2)

tensorflow PYPI version =2.3.0, =1.0.62, =1.2.0, =0.0.1a0, =0.0.1, =1.0.0rc1, =20210206.0.0, =0.1.0.dev1, =0.2.4, =1.0.1.0, =0.1.0, =2.0.0rc1, =2.0.3 and more Source cves: CVE-2021-37646 Source advisory: OSV:PYSEC-2021-268...

Sharing the first SimuLand dataset to expedite research and learn about adversary tradecraft

Last month, we introduced the SimuLand project to help security researchers around the world deploy lab environments to reproduce well-known attack scenarios, actively test detections, and learn more about the underlying behavior and implementation of adversary techniques. Since the release of th...

Awesome-Red-Teaming

This is an offensive tool for Red Teaming. It is a list of resources for anyone wishing to learn about Red Teaming, including initial access, execution, persistence, privilege escalation, defense evasion, credential access, discovery, lateral movement, collection, exfiltration, command and contro...

OpenAttack - An Open-Source Package For Textual Adversarial Attack

OpenAttack is an open-source Python-based textual adversarial attack toolkit, which handles the whole process of textual adversarial attacking, including preprocessing text, accessing the victim model, generating adversarial examples and evaluation. Features & Uses OpenAttack has following...

WAF-A-MoLE - A Guided Mutation-Based Fuzzer For ML-based Web Application Firewalls

A guided mutation-based fuzzer for ML-based Web Application Firewalls, inspired by AFL and based on the FuzzingBook by Andreas Zeller et al. Given an input SQL injection query, it tries to produce a semantic invariant query that is able to bypass the target WAF. You can use this tool for assessin...

How to confuse antimalware neural networks. Adversarial attacks and protection

Introduction Nowadays, cybersecurity companies implement a variety of methods to discover new, previously unknown malware files. Machine learning ML is a powerful and widely used approach for this task. At Kaspersky we have a number of complex ML models based on different file features, including...

adversarial-robustness-toolbox (1.2.0), aristote-malo (1.9.5) +79 more potentially affected by CVE-2021-28677 via pillow (7.0.0)

pillow PYPI version 7.0.0 is affected by a known vulnerability. The following packages have a transitive dependency on pillow and may be impacted: - adversarial-robustness-toolbox 1.2.0 - aristote-malo 1.9.5 - auto-annotate =1.0.0, =1.0.2, =0.1.0.dev31, =1.0.1, =0.2.3, =0.0.1.25, =0.2.1, =0.0.2,...