TLDR InfoSec

TLDR INFORMATION SECURITY 2025-03-17

6 sections19 entrieslinks blockcomplete parse

Parse notes

  • Subscription/account footer truncated before parsefooter_truncated

01GENERAL

Sponsored

REVIEWS AND CONTENT MANAGEMENT

Is outdated and messy documentation slowing down your security, compliance, and privacy reviews? It's time for a reset.

safebase.ioSponsor

02ATTACKS & VULNERABILITIES

CLICKFIX: HOW TO INFECT YOUR PC IN THREE EASY STEPS

The "ClickFix" scheme tricks users into downloading password-stealing malware disguised as CAPTCHA verification. Users press Windows+R, CTRL+V, and Enter, which activates the download through mshta.exe. Organizations can prevent this by disabling the run command for the Windows+R shortcut.

krebsonsecurity.com3 min read

A DOGE STAFFER BROKE TREASURY POLICY BY EMAILING UNENCRYPTED PERSONAL DATA

A 25-year-old DOGE staffer violated Treasury policy by sending an unencrypted spreadsheet with personal data to Trump officials. Marko Elez, who previously worked for Musk's companies, had been "mistakenly" granted access to Treasury systems before resigning in February due to racist posts. He was rehired at the Social Security Administration despite ongoing litigation.

theverge.com2 min read

GITHUB ACTION TJ-ACTIONS/CHANGED-FILES SUPPLY CHAIN ATTACK: EVERYTHING YOU NEED TO KNOW

The tj-actions/changed-files GitHub Action repository was compromised by an attacker posing as the Renovate bot. The attacker added a malicious payload that caused the CI runner to dump its memory into the Actions work log and redirected all tags to include this payload. Sensitive data in GitHub secrets for this action can be visible in public repos, but the impact on private repositories is more limited.

wiz.io5 min read

03STRATEGIES & TACTICS

GUARD YOUR CODEBASE: PRACTICAL STEPS AND TOOLS TO PREVENT MALICIOUS CODE

Apiiro researchers found that targeting specific coding anti-patterns enabled them to detect malicious packages with a 94.3% success rate in Python and 88.4% in npm. They focused on obfuscated code and dynamic code execution and have released Semgrep rules for detection.

apiiro.com10 min read

HIERARCHY OF NEEDS

Swannman's Hierarchy of Needs Pyramid model outlines 10 essential layers for establishing effective security functions, with threat hunting as a specialized function at the top. The Hierarchy of Needs aids organizations in planning and assessing incident response capabilities by identifying advanced metrics for complex questions. The Plateaus Model tracks organizational maturity throughout the hierarchy.

huntbook.predefender.com7 min read

WHOAMI - GET PRINCIPAL NAME FROM KEYS

Attackers often use get-caller-identity to find IAM roles linked to keys, but this API call can be monitored. This post shares a tactic to bypass these detections by querying an API that returns caller identity on a failed call.

hackingthe.cloud2 min read

04LAUNCHES & TOOLS

TUMERYK (PRODUCT LAUNCH)

Tumeryk's AI Trust Scores assesses risk using nine factors, offering CISOs objective security scores for gen AI models.

tumeryk.com

YAAK API CLIENT

Yaak is a desktop client for interacting with REST, GraphQL, WebSocket, and gRPC APIs.

github.comGitHub repository

NULLGATE

NullGate provides a modern approach to using NTAPI functions via indirect syscalls and the FreshyCalls method for dynamic syscall number retrieval. It includes a novel technique to bypass Windows Defender's memory scanning and features a classic PoC process injector.

github.comGitHub repository

05MISCELLANEOUS

DEMS ASK FEDERAL AGENCIES FOR REASSURANCE DOGE ISN'T FEEDING DATA INTO AI WILLY-NILLY

House Democrats have questioned 24 federal agencies about Elon Musk's DOGE team possibly sharing sensitive government data with unauthorized AI systems. Rep. Gerald Connolly raised concerns about legal violations, data privacy, and handling of Americans' personal information, requesting responses on DOGE's AI activities by March 26.

theregister.com3 min read

FAKE "SECURITY ALERT" ISSUES ON GITHUB USE OAUTH APP TO HIJACK ACCOUNTS

A phishing campaign on GitHub uses fake "Security Alert" issues to trick users into granting access to a malicious OAuth app, giving attackers complete control over accounts and code. The alerts mention unusual activity from Reykjavik, Iceland, and the 53.253.117.8 IP address, urging users to allow risky permissions. Affected users must revoke access, inspect for suspicious apps, and change credentials immediately.

bleepingcomputer.com2 min read