TLDR InfoSec

TLDR INFORMATION SECURITY 2025-09-22

6 sections20 entrieslinks blockcomplete parse

Parse notes

  • Subscription/account footer truncated before parsefooter_truncated

01GENERAL

Sponsored

WHAT COMES AFTER THE CORPORATE FIREWALL?

The corporate firewall is a relic of bygone times. With employees working from coffee shops, airports, and home networks, traditional perimeter security leaves your endpoints exposed to unmanaged devices and compromised networks.

threatlocker.comSponsor

02ATTACKS & VULNERABILITIES

FORTRA ADDRESSED A MAXIMUM SEVERITY FLAW IN GOANYWHERE MFT SOFTWARE

Fortra has patched a critical deserialization vulnerability (CVE-2025-10035) with a maximum CVSS score of 10.0 in the GoAnywhere Managed File Transfer software that allows attackers to execute arbitrary commands through the License Servlet. The company has urged customers to upgrade to version 7.8.4 or sustaining release 7.6.3 and recommends restricting public access to the GoAnywhere Admin Console to mitigate exploitation risks. This follows previous high-severity vulnerabilities in the same product, including an authentication bypass flaw (CVE-2024-0204) discovered in January 2024.

securityaffairs.com2 min read

ATTACKERS ABUSE AI TOOLS TO GENERATE FAKE CAPTCHAS IN PHISHING ATTACKS

Trend Micro researchers have discovered a new phishing campaign that utilizes Lovable, Netlify, and Vercel to build and host vibe-coded phishing campaigns posing as CAPTCHA pages. The campaigns begin with an email carrying urgent messages such as “Password Reset Required,” which directs the user to a fake CAPTCHA. The fake CAPTCHA then redirects the user to the phishing page, which harvests their credentials. The CAPTCHA pages make the attacks seem more legitimate and filter out security tools.

infosecurity-magazine.com2 min read

NEW YORK BLOOD CENTER ALERTS 194K PEOPLE TO DATA BREACH

The New York Blood Center (NYBCe) has confirmed that nearly 194K people had their data breached during an attack from January 20 to January 26. The breached data includes names, SSNs, driver's license, state ID numbers, bank account details, health information, and test results.

infosecurity-magazine.com2 min read

03STRATEGIES & TACTICS

YOU DON'T NEED QUANTUM HARDWARE FOR POST-QUANTUM SECURITY

Organizations can prepare for quantum computing threats (Q-day) using post-quantum cryptography (PQC) deployed on existing hardware without needing expensive quantum technologies like quantum key distribution (QKD) or quantum random number generators (QRNG). Being “quantum ready” means systems remain secure after quantum computers can break conventional cryptography. Beware of vendor claims that quantum hardware products are necessary for quantum-resistant security. Organizations should prioritize implementing PQC algorithms on current infrastructure rather than investing in quantum hardware solutions that are neither necessary nor sufficient for protection against quantum adversaries.

blog.cloudflare.com15 min read

HOW TO MANAGE AI BOTS WITH AWS WAF AND ADVANCED SECURITY

AI bots can be problematic as they can use proprietary data to train models, cause performance degradation, or allow for unwanted agentic/automatic behavior. Some bots, such as Applebot, Amazonbot, and Googlebot, respect robots.txt and will obey directives that tell them not to train on a website's data. AWS WAF can offer rate-limiting rules, CAPTCHA, or honeypots for the Bot Control group to account for bots that don't respect robots.txt.

aws.amazon.com5 min read

MONITORING BASH `READLINE` WITH EBPF CO-RE & LLM ASSISTANCE

Although the eBPF Compiler Collection (BCC) can monitor bash commands, it compiles programs dynamically, so the complete Clang/LLVM toolchain must be shipped to every machine. eBPF CO-RE (Compile Once - Run Everywhere) can enable portability across different kernel versions without recompilation. This article walks through how to write an eBPF program to monitor the `readline` program to monitor bash commands.

medium.com7 min read

04LAUNCHES & TOOLS

Sponsored

IS YOUR THREAT INTELLIGENCE MISSING THE ACTUAL CONVERSATION?

Static threat feeds only skim the surface. While you're reading yesterday's news, threat actors are coordinating in closed Telegram groups and private forums that are rarely reached by threat intelligence vendors. Primary Source Collection [12] gets you inside those spaces - with real-time access to encrypted chats, invite-only communities, and adversarial planning. Read the Flashpoint whitepaper

go.flashpoint.ioSponsor

go.flashpoint.io

BYTECASTER

ByteCaster is a new tool that aims to be a Swiss Army Knife for payload encryption, obfuscation, and conversion to byte arrays in a single command. It supports three encryption algorithms, four encoding/obfuscation algorithms, and 14 output formats.

github.comGitHub repository

HAVE I BEEN PWNED DEMOS ARE NOW LIVE

Troy Hunt has launched a new series of demos exploring different features of the Have I Been Pwned platform. The first demo covers domain searches.

troyhunt.com2 min read

SPEARSPRAY

SprearSpray is an advanced password spraying tool for Active Directory environments. It combines LDAP user enumeration with pattern-based password generation.

github.comGitHub repository

05MISCELLANEOUS

DISRUPTION AT EUROPEAN AIRPORTS EXPECTED TO CONTINUE THROUGH SUNDAY

A cyberattack on Collins Aerospace's check-in and boarding systems disrupted operations at major European airports, including Heathrow, Brussels, and Berlin, forcing airports to rely on manual check-in procedures and causing flight delays and cancellations that continued into Sunday. Brussels Airport asked airlines to cancel half of their Sunday departing flights to mitigate disruptions. 29 total departures and arrivals were cancelled across the three main affected airports. The European Commission confirmed the incident is under investigation, though officials stated there's no evidence of a widespread or severe attack.

links.tldrnewsletter.com2 min read

CHINESE NETWORK SELLING THOUSANDS OF FAKE US AND CANADIAN IDS

CloudSEK exposed a China-based operation called “ForgeCraft” that sold over 6,500 counterfeit US and Canadian driver's licenses and Social Security cards to more than 4,500 buyers, generating over $785,000 in revenue through 83 websites and using covert packaging methods to ship fake IDs hidden in everyday items. The sophisticated fake documents feature scannable barcodes, holograms, and UV markings and pose national security risks by enabling financial fraud, bypassing border checks, and potentially facilitating voter fraud. CloudSEK researchers identified the main operator's location in Xiamen, China, and shared evidence with authorities to disrupt the operation.

hackread.com3 min read