TLDR InfoSec

TLDR INFORMATION SECURITY 2024-04-08

6 sections17 entrieslinks blockcomplete parse

Parse notes

  • Subscription/account footer truncated before parsefooter_truncated

01GENERAL

Sponsored

YOU DO ABOUT IT?

Credentials serve as the keys to our online existence — and are also a security liability. The average professional manages up to 200 sets of credentials [3] on average; when they’re compromised, cyber attackers gain instant entry into sensitive systems and can often move laterally to find your crown jewels.

lacework.comSponsor

02ATTACKS & VULNERABILITIES

HOYA’S OPTICS PRODUCTION AND ORDERS DISRUPTED BY CYBERATTACK

Japanese optical product manufacturer Hoya Corporation announced that it is working to restore systems at some production plants following a cyberattack. The incident prompted Hoya to isolate servers, impacting IT systems at its headquarters and business divisions. The company has engaged forensic investigators and notified relevant authorities as it investigates and responds to the attack, which caused system failures across its operations.

bleepingcomputer.com2 min read

MISSOURI COUNTY DECLARES STATE OF EMERGENCY AMID SUSPECTED RANSOMWARE ATTACK

Jackson County, Missouri, has declared a state of emergency and indefinitely closed key offices like Assessment, Collection, and Recorder of Deeds in response to a suspected ransomware attack that disrupted its IT systems. The attack has rendered certain systems inoperable, impacting services like tax payments, marriage licenses, and inmate searches, while other systems remain functional.

arstechnica.com3 min read

ANTHROPIC RESEARCHERS WEAR DOWN AI ETHICS WITH REPEATED QUESTIONS

Researchers at Anthropic have devised a new technique to convince AI models to answer questions that they aren’t meant to. The method involves exploiting larger context windows to prime the LLM with a series of questions (that it won’t answer due to filters) and answers as if the LLM answered them. Over time, the LLM will get “confused” and think it answered the questions and will answer inappropriate questions.

techcrunch.com3 min read

03STRATEGIES & TACTICS

BYPASSING ANTI-REVERSING DEFENSES IN IOS APPLICATIONS

This blog post explains how to bypass anti-reversing defenses, such as ptrace syscall and getppid() method, in iOS applications. It also covers bypassing sysctl syscall and jailbreak detection to analyze and manipulate application behavior at runtime. The author demonstrates methods to overcome these security measures using tools like frida and radare2.

twelvesec.com12 min read

APPLYING LLMS TO THREAT INTELLIGENCE

This article discusses the use of LLMs to enhance the capabilities of threat intelligence analysts. It explains the technique of prompt engineering and provides an example of an ideally crafted prompt. The article also explores the practical applications of LLMs, few-shot prompting, and retrieval augmented generation (RAG).

blog.securitybreak.io12 min read

04LAUNCHES & TOOLS

BINARLY (PRODUCT LAUNCH)

Binarly is a firmware supply chain security platform that uses AI to detect and remediate known and unknown vulnerabilities at every step of the device and software supply chain.

binarly.io

INSECURE KUBERNETES DEPLOYMENTS

Test every type of Container configuration scanner against a single repo that's comically insecure with documented issues

github.comGitHub repository

C2-TRACKER

A live feed of C2 servers, tools, and botnets. The tool scrapes Shodan for IoCs of various malicious activities.

github.comGitHub repository

05MISCELLANEOUS

NIST GRANTS $3.6 MILLION TO BOOST US CYBERSECURITY WORKFORCE

NIST has announced $3.6 million in grants to 18 education and community organizations across 15 states to address the cybersecurity skills shortage. Each state will receive around $200,000 to educate future cybersecurity employees through Regional Alliances and Multistakeholder Partnerships to Stimulate (RAMPS) projects aligned with local industry needs, overseen by the NICE cybersecurity workforce development partnership.

securityweek.com3 min read

WIZ RESEARCH FINDS ARCHITECTURE RISKS THAT MAY COMPROMISE AI-AS-A-SERVICE PROVIDERS AND CONSEQUENTLY RISK CUSTOMER DATA; WORKS WITH HUGGING FACE ON MITIGATIONS

Critical risks in AI-as-a-service providers that could allow threat actors to escalate privileges, gain cross-tenant access to other customers' models, and more, have been identified. These risks stem from the ability to run malicious models uploaded in pickle format on shared inference infrastructure and exploit shared CI/CD systems.

wiz.io12 min read

WHY THE WORLD NEEDS A NEW CYBER TREATY FOR CRITICAL INFRASTRUCTURE

With attacks on critical infrastructure in the energy and healthcare sectors continuously increasing, it’s time for countries to adopt binding international legislation that builds a new cyber treaty. There have been some efforts made, but a coordinated response could solidify previous efforts while building accountability and allowing for resource allocation towards protecting critical infrastructure.

carnegieeurope.eu15 min read