TLDR InfoSec

TLDR INFORMATION SECURITY 2024-05-29

5 sections17 entrieslinks blockcomplete parse

Parse notes

  • Subscription/account footer truncated before parsefooter_truncated

01ATTACKS & VULNERABILITIES

2.8 MILLION IMPACTED BY DATA BREACH AT PRESCRIPTION SERVICES FIRM SAV-RX

Pharmacy prescription services provider Sav-Rx is notifying around 2.8 million individuals that their personal information was compromised in a cyberattack on October 8, 2023. The compromised information includes names, addresses, dates of birth, email addresses, phone numbers, Social Security numbers, eligibility data, and insurance identification numbers.

securityweek.com3 min read

INTERNET ARCHIVE AND THE WAYBACK MACHINE UNDER DDOS CYBER-ATTACK

The Internet Archive, a nonprofit digital library housing millions of historical documents and websites, has endured an intermittent Distributed Denial-of-Service (DDoS) cyber-attack for three days, limiting access to its vast collections including the Wayback Machine's web archive.

blog.archive.org2 min read

TP-LINK FIXES CRITICAL RCE IN POPULAR GAMING ROUTER

An RCE vulnerability TP-Link Archer C5400X router was recently disclosed after identification by OneKey. The vulnerability is a command injection and buffer overflow in the rftest binary, which is responsible for network self assessments. A patch that implements input sanitization was released to address the vulnerability.

bleepingcomputer.com2 min read

02STRATEGIES & TACTICS

BEST PENETRATION TESTING TOOLS IN 2024

Penetration testing tools are essential for efficient hacking and security assessments. While hacking without tools is possible, it is time-consuming and repetitive. These tools streamline basic scanning and security processes, saving time and effort. This post dives into some of the best penetration testing tools.

medium.com8 min read

STOP RECOMMENDING JWTS (WITH SYMMETRIC KEYS)

The Truffle Security team analyzed 10k JWTs found via Shodan and another 500 committed to public repos in GitHub. They used hashcat to attempt cracking the JWTs and found that 1.2% of symmetric keys used in production were guessable within seconds. This was caused by users implementing weak passwords (including the defaults used in documentation and examples) as their symmetric keys when setting up their authentication systems.

trufflesecurity.com5 min read

A BASIC GUIDE TO DISCOVERING ATTACK SURFACE WITH GHIDRA AND GDB

This blog post provides a walkthrough on how to programmatically map the attack surface of an application using GDB and Ghidra. It begins with walking through the strategy for identifying functions after ASLR and then shows how it can be implemented using GDB Python. The post takes this one step further and shows how to write a Ghidra Python script to generate the GDB Python script for each function and ignore artificially generated blocks.

medium.com5 min read

03LAUNCHES & TOOLS

SNAP WTF MACOS

WTF Snapshot fuzzing of macOS targets, enabling users to perform precisely targeted fuzz testing of hard-to-pinpoint chunks of the macOS kernel.

github.comGitHub repository

ARCJET JS

Arcjet JS SDKs. Rate limiting, bot protection, email verification, and attack defense for Node.js, Next.js, Bun, and SvelteKit.

github.comGitHub repository

SUDO FOR WINDOWS

Sudo for Windows is a windows-specific implementation of the Sudo concept to escalate privilege from Linux/Unix. This means that certain elements of the traditional sudo experience are not present in Sudo for Windows and vice versa. Scripts and documentation that are written for sudo may not be able to be used directly with Sudo for Windows without some modification.

github.comGitHub repository

04MISCELLANEOUS

OPENAI'S NEW SAFETY COMMITTEE IS MADE UP OF ALL INSIDERS

OpenAI has formed an internal Safety and Security Committee headed by CEO Sam Altman and composed of company insiders. Its goal is to oversee critical security decisions related to OpenAI's AI projects over the next 90 days, raising concerns among ethicists about the lack of independent oversight.

techcrunch.com3 min read

AI RED TEAMING TOOLS HELPED X-FORCE BREAK INTO A TECH MANUFACTURER IN 8 HOURS

IBM reported that its X-Force red team was able to compromise a network that they had allocated 3 weeks for in just 8 hours with the help of AI red teaming tools. The team found a vulnerability in the company's HR portal with the aid of the tooling and then installed a rootkit after exploiting the vulnerability and eventually pivoted through the network. IBM believes that these AI tools can be essential in analyzing the large volume of data accumulated in a red team assessment.

theregister.com3 min read