TLDR InfoSec

TLDR INFORMATION SECURITY 2025-03-13

6 sections18 entrieslinks blockcomplete parse

Parse notes

  • Subscription/account footer truncated before parsefooter_truncated

01GENERAL

02ATTACKS & VULNERABILITIES

POWERSCHOOL PORTAL COMPROMISED MONTHS BEFORE MASSIVE DATA BREACH

Hackers accessed PowerSchool's support portal months before a December 2024 data breach, potentially impacting 70 million individuals. They used compromised credentials to steal personal data, including names, contact details, birth dates, medical information, and Social Security numbers.

securityweek.com2 min read

PREVIOUSLY UNIDENTIFIED BOTNET TARGETS UNPATCHED TP-LINK ARCHER HOME ROUTERS

Cato Networks has detected a new botnet named Ballista that is targeting unpatched TP-Link Archer routers using the CVE-2023-1389 exploit. Researchers suggest the hacker is likely from Italy and that the botnet is in its early stages, with evolving malware capabilities. Unlike other DDoS-for-hire botnets, Ballista's malware can also execute arbitrary commands on remote systems.

therecord.media3 min read

03STRATEGIES & TACTICS

AZURE'S WEAKEST LINK? HOW API CONNECTIONS SPILL SECRETS

Binary Security found undocumented Azure API Connections enabling privilege escalation. With Reader permissions, attackers can GET requests to backend resources like Key Vaults, Defender ATP, and third-party services such as Jira and Salesforce.

binsec.no15 min read

COPYOBJECTION: FENDING OFF RANSOMWARE IN AWS

Attackers are using SSE-C to encrypt objects in ransomware attacks on AWS accounts. Jesse Griggs established a test environment to measure the alert response time through CloudWatch and how quickly Lambda could block the attacker or disable SSE-C. He noted alerts took about 6 minutes to arrive, while automated responses occurred immediately.

redcanary.com6 min read

BEYOND THE HOOK: A TECHNICAL DEEP DIVE INTO MODERN PHISHING METHODOLOGIES

This article explores modern phishing techniques, from basic HTML pages to those that bypass multi-factor authentication. It covers attack methods such as Browser-in-the-Browser (BITB), Attacker-in-the-Middle (AITM), and WebRTC, as well as infrastructure considerations for hosting and delivery. Phishing has evolved with sophisticated techniques using cloud provider domains, making detection harder. Effective defense requires multi-layered approaches combining technical controls and human awareness.

blog.quarkslab.com31 min read

04LAUNCHES & TOOLS

DEVCONTAINER

devcontainer always tries to find the best way to install the most popular tools, so they can all work seamlessly, and at the same time add security by default.

github.comGitHub repository

FAST INCIDENT RESPONSE

Fast Incident Response (FIR) is a cybersecurity IR management platform from the French Cert Societe Generale.

github.comGitHub repository

STREETCRED

StreetCred is a tool created for Red Team operators to test default credentials and execute scripts over SSH and WinRM.

github.comGitHub repository

05MISCELLANEOUS

CISA WORKER SAYS 100-STRONG RED TEAM FIRED AFTER DOGE CANCELLED CONTRACT

CISA's 100-member penetration testing team was terminated after Trump's DOGE unit canceled their contract, leading to cuts in another red team. The election security group EI-ISAC also shut down due to funding loss, raising concerns about cyberattack vulnerabilities.

theregister.com3 min read

IN-DEPTH TECHNICAL ANALYSIS OF THE BYBIT HACK

Bybit experienced the largest cryptocurrency theft ever, totaling $1.4 billion, when attackers compromised a developer machine and injected malicious JavaScript into Safe{Wallet} UI, altering what signers saw. This off-chain attack tricked signers into approving a delegatecall that replaced the proxy implementation with the attackers' code, giving them control over all funds. The Lazarus group from North Korea executed this sophisticated attack by exploiting human verification vulnerabilities and the lack of human-readable formats for complex smart contract operations in the EIP-712 standard.

nccgroup.com11 min read

RESEARCH UNVEILS THAT MORE THAN 110,000 IOS APPS LEAK SECRETS

Over 110,000 iOS apps leak secrets, including sensitive data like cloud storage keys. Most apps expose at least one hardcoded secret, risking user breaches. Apple's App Store does not check for these vulnerabilities, underscoring the need for better security practices.

links.tldrnewsletter.com8 min read