Daily Edition · Page 5

THE DAILY INDEX

Technology, AI, software, security, crypto and business

ATTACKS & VULNERABILITIES

TLDR InfoSec

SALT TYPHOON USED CUSTOM MALWARE JUMBLEDPATH TO SPY U.S. TELECOM PROVIDERS

Cisco's Talos group has noted that China-linked cyber espionage group Salt Typhoon used custom malware JumbledPath to spy on U.S. telecom providers, breaching major firms for over three years. The group exploited vulnerabilities in Cisco devices, stole credentials, and manipulated network settings to gain unauthorized access. The attacks targeted multiple countries' telecom providers, prompting security measures from companies like Lumen, AT&T, and Verizon.

FIND HIDDEN AWS RESOURCES WITH EFFECTIVE WORDLISTS

This post proposes a method for generating wordlists to enumerate resources in a target AWS account using Awseye. The process involves exporting mentions from Awseye, normalizing the account IDs and regions, and then counting and sorting the results to see the most commonly mentioned resources. The post highlights some alternatives such as parsing AWS documentation, querying Sourcegraph, or using AI-generated lists.

AN INSIDE LOOK AT NSA (EQUATION GROUP) TTPS FROM CHINA'S LENSE

Reports coming out of Chinese cybersecurity entities like Qihoo 360, Pangu Lab, and the National Computer Virus Emergency Response Center (CVERC) attributed cyber operations to the NSA's APT-C-40 group. The article goes into detail about the attribution methodology used as well as identifying tools found before the Shadow Broker's leak. The investigation revealed TAO's use of sophisticated tools like SECONDDATE and NOPEN to conduct remote control, data exfiltration, and credential harvesting.

AGENTIC AI THREAT MODELING FRAMEWORK: MAESTRO

MAESTRO is a threat modeling framework that was built to address gaps in traditional threat modeling frameworks like STRIDE and PASTA related to the unique challenges that are inherent in AI systems. It is designed for multi-agent environments and is based on a seven-layer reference architecture that spans from the agent ecosystem at the top to foundational models at the bottom. The model also provides threat modeling examples of agentic architecture patterns such as single agent, multi agent, unconstrained conversational autonomy, human in the loop, and task-oriented agents.

TLDR IndexPage 5