Huawei MindSpore Community numeric error vulnerability

Huawei MindSpore Community is an open source deep learning framework from Huawei China.A numerical error vulnerability exists in versions prior to Huawei MindSpore Community 1.3.0, which stems from the fact that when performing the initialization operation of the Split operator, if a dimension in...

NVIDIA DGX 缓冲区错误漏洞

NVIDIA DGX is a high-performance workstation for deep learning applications from NVIDIA. The NVIDIA DGX A100 suffers from a buffer error vulnerability that originates from accessing an uninitialized pointer to SBIOS in Ofbd. An attacker could exploit this vulnerability to execute arbitrary code o...

NVIDIA DGX 输入验证错误漏洞

NVIDIA DGX is a high-performance workstation for deep learning applications from NVIDIA. An input validation error vulnerability exists in NVIDIA DGX A100, which stems from incorrect validation of the SBIOS array index in IpSecDxe. An attacker could exploit this vulnerability to execute arbitrary...

Security Tool Guts: How Much Should Customers See?

Many cybersecurity tools use engines that calculate risk for events in customer environments. The accuracy of these risk engines is a major concern for customers, since it determines whether an attack is detected or not. Therefore, organizations often request visibility into how a risk engine...

Combing through the fuzz: Using fuzzy hashing and deep learning to counter malware detection evasion techniques

Today’s cybersecurity threats continue to find ways to fly and stay under the radar. Cybercriminals use polymorphic malware because a slight change in the binary code or script could allow the said threats to avoid detection by traditional antivirus software. Threat actors customize their wares...

Combing through the fuzz: Using fuzzy hashing and deep learning to counter malware detection evasion techniques

Today’s cybersecurity threats continue to find ways to fly and stay under the radar. Cybercriminals use polymorphic malware because a slight change in the binary code or script could allow the said threats to avoid detection by traditional antivirus software. Threat actors customize their wares...

In0ri - Defacement Detection With Deep Learning

In0ri is a defacement detection system utilizing a image-classification convolutional neural network. Introduction When monitoring a website, In0ri will periodically take a screenshot of the website then put it through a preprocessor that will resize the image down to 250x250px and numericalize t...

Microsoft announces recipients of academic grants for AI research on combating phishing

Every day in the ever-changing technology landscape, we see boundaries shift as new ideas challenge the old status quo. This constant shift is observed in the increasingly sophisticated and connected tools, products, and services people and organizations use on a daily basis, but also in the...

Microsoft announces recipients of academic grants for AI research on combating phishing

Every day in the ever-changing technology landscape, we see boundaries shift as new ideas challenge the old status quo. This constant shift is observed in the increasingly sophisticated and connected tools, products, and services people and organizations use on a daily basis, but also in the...

qianjunakasumi kongchuanhujiao 授权问题漏洞

qianjunakasumi kongchuanhujiao is qianjunakasumi an open source application . An online teaching quiz statistics deep learning analytics system . A security vulnerability exists in github.com/kongchuanhujiao/server before version 1.3.21, which stems from an authentication bypass...

Code Injection in jeikeilim/kindle

Description Kindle is an easy model build package for PyTorch. Building a deep learning model became so simple that almost all model can be made by copy and paste from other existing model codes, which is vulnerable to Arbitary Code Execution. Vulnerability Vulnerable to YAML deserialization atta...

Training Transformers for Cyber Security Tasks: A Case Study on Malicious URL Prediction

Highlights Perform a case study on using Transformer models to solve cyber security problems Train a Transformer model to detect malicious URLs under multiple training regimes Compare our model against other deep learning methods, and show it performs on-par with other top-scoring models Identify...

in catalyst-team/catalyst

Description Catalyst is a PyTorch framework for Deep Learning research and development. It focuses on reproducibility, rapid experimentation, and codebase reuse so you can create something new rather than write another regular train loop. This package was vulnerable to Arbitrary code execution vi...

in nvidia/runx

Description runx is a Deep Learning Experiment Management library by NVIDIA. This package was vulnerable to Arbitrary code execution via Insecure YAML deserialization due to the use of a known vulnerable function load in yaml. repo: https://github.com/NVIDIA/runx Proof of Concept python...

SoReL-20M: A Huge Dataset of 20 Million Malware Samples Released Online

Cybersecurity firms Sophos and ReversingLabs on Monday jointly released the first-ever production-scale malware research dataset to be made available to the general public that aims to build effective defenses and drive industry-wide improvements in security detection and response. "SoReL-20M"...

Security Unlocked—A new podcast exploring the people and AI that power Microsoft Security solutions

It’s hard to keep pace with all the changes happening in the world of cybersecurity. Security experts and leaders must continue learning and unlearning to stay ahead of the ever-evolving threat landscape. In fact, many of us are in this field because of our desire to continuously challenge...

Detecting Deep Fakes with a Heartbeat

Researchers can detect deep fakes because they dont convincingly mimic human blood circulation in the face: In particular, video of a persons face contains subtle shifts in color that result from pulses in blood circulation. You might imagine that these changes would be too minute to detect merel...

Seeing the big picture: Deep learning-based fusion of behavior signals for threat detection

The application of deep learning and other machine learning methods to threat detection on endpoints, email and docs, apps, and identities drives a significant piece of the coordinated defense delivered by Microsoft Threat Protection. Within each domain as well as across domains, machine learning...

Microsoft researchers work with Intel Labs to explore new deep learning approaches for malware classification

The opportunities for innovative approaches to threat detection through deep learning, a category of algorithms within the larger framework of machine learning, are vast. Microsoft Threat Protection today uses multiple deep learning-based classifiers that detect advanced threats, for example,...

Top 10 Most Innovative Cybersecurity Companies After RSA 2020

The RSA Conference, the world's leading information security conference and exposition, held its 29th annual event in San Francisco last week. According to the organizers, over 36,000 attendees, 704 speakers, and 658 exhibitors gathered at the Moscone Center to discuss privacy, Machine Learning,...