TLDR InfoSec

TLDR INFORMATION SECURITY 2024-05-20

5 sections17 entrieslinks blockcomplete parse

Parse notes

  • Subscription/account footer truncated before parsefooter_truncated

01ATTACKS & VULNERABILITIES

AUSSIE COPS PROBE MEDISECURE'S 'LARGE-SCALE RANSOMWARE DATA BREACH'

MediSecure, an Australian e-prescription provider, has fallen victim to a ransomware attack that resulted in the suspected theft of patients' personal and health data. The company stated that the incident likely originated from a third-party vendor breach. MediSecure has notified regulators about the large-scale data breach, which impacted an unspecified number of individuals.

theregister.com3 min read

CRITICAL FLAW IN AI PYTHON PACKAGE CAN LEAD TO SYSTEM AND DATA COMPROMISE

A critical vulnerability dubbed "Llama Drama" (CVE-2024-34359) has been discovered in the llama_cpp_python package used for integrating AI models with Python. The flaw, related to the improper use of the Jinja2 template rendering tool, enables template injection attacks and allows arbitrary code execution, putting AI application systems and data at risk.

securityweek.com3 min read

02STRATEGIES & TACTICS

THREAT DETECTION ENGINEERING AND INCIDENT RESPONSE WITH AUDITD AND SENTINEL

This article explains how to use AuditD to detect threats by creating rules based on logs and analyzing them further. It provides a simple example rule to monitor executions in the /tmp directory, emphasizing the importance of monitoring to enhance security. The article offers insights into investigating AuditD events and conducting in-depth analysis for security monitoring.

bit.ly3 min read

RELATIVE PATH FILE INJECTION: THE NEXT EVOLUTION IN RPO

This article introduces a new attack method for injecting rogue files and forcing victims to download them. The attack requires a page to have a path confusion vulnerability, a download link with an anchor tag that uses relative paths, and content that is injectable via a persistent XSS or XSS cache poisoning. The attack expands on RPO because it doesn't rely on deprecated technologies.

blog.ionatomics.org7 min read

03LAUNCHES & TOOLS

SYDELABS

Sydeabs is a security and risk management platform aimed at securing Generative AI throughout the entire development lifecycle.

sydelabs.aiTool

JS-X-RAY

JavaScript & Node.js open-source SAST scanner. A static analyzer for detecting most common malicious patterns.

github.comGitHub repository

REVERST

Reverst is a load-balanced, Golang reverse tunnel that runs over HTTP3/QUIC. The tunnel server is meant to be deployed publicly while clients in restrictive networks connect out to it.

github.comGitHub repository

04MISCELLANEOUS

YARA IS DEAD, LONG LIVE YARA-X

YARA has been rewritten as YARA-X in Rust for better user experience, rule-level compatibility, performance, reliability, and developer-friendliness. The rewrite was necessary due to significant design changes and the limitations of the existing C codebase. YARA-X is currently stable but in beta. The project aims to surpass YARA in all aspects through ongoing enhancements.

virustotal.github.io3 min read

SLACK UNDER ATTACK OVER SNEAKY AI TRAINING POLICY

Slack is facing backlash from users over its approach to using customer data for training AI models. While many tech companies leverage user data for AI training, Slack's confusing and seemingly outdated privacy policy allows it to opt users in by default, requiring them to actively email the company to opt out.

techcrunch.com4 min read