TLDR InfoSec

TLDR INFORMATION SECURITY 2024-05-01

5 sections13 entrieslinks blockcomplete parse

Parse notes

  • Subscription/account footer truncated before parsefooter_truncated

01ATTACKS & VULNERABILITIES

GOOGLE DEFENDED THE PLAY STORE FROM 2.28 MILLION MALICIOUS APPS LAST YEAR

Google's enhanced security measures helped block 2.28 million privacy-violating apps from its Play Store in 2023. The company banned 333,000 bad accounts, rejected nearly 200,000 apps using sensitive permissions, and worked with SDK providers to reduce data access and sharing, impacting over 790,000 apps and improving user privacy.

androidpolice.com4 min read

VERIZON, AT&T, T-MOBILE, AND SPRINT FINED ALMOST $200M FOR SHARING CUSTOMER DATA

The FCC reported that Verizon, AT&T, T-Mobile, and Sprint sold data “aggregators” access to their customers' location data. This data was then resold to third-party location-based service providers. The FCC has issued a fine for circumventing the requirement to obtain user consent after the telecommunications providers were alerted of this issue and did not remediate it.

qz.com3 min read

PHILADELPHIA INQUIRER: DATA OF OVER 25,000 PEOPLE STOLEN IN 2023 BREACH

Over 25,000 people had their personal and financial data stolen in a breach at the Philadelphia Inquirer in May 2023. The breach disrupted newspaper operations, but the website remained unaffected. The ransomware group responsible, named Cuba, published stolen files after the newspaper refused to pay a ransom.

bleepingcomputer.com2 min read

02STRATEGIES & TACTICS

NO, LLM AGENTS CAN NOT AUTONOMOUSLY EXPLOIT ONE-DAY VULNERABILITIES

This blog post takes a critical look at an academic paper that received widespread media attention which claimed to have built an LLM Agent that can exploit one-day vulnerabilities. The author points out that the paper included an agent that had Internet access and that the authors of the paper chose CVEs that had very detailed exploits and PoCs available. The post concludes that the paper demonstrated the capabilities of GPT-4 as an intelligent scanner and crawler but based on the data provided didn't demonstrate its ability to rediscover these vulnerabilities or generate novel exploit code.

struct.github.io4 min read

03LAUNCHES & TOOLS

SPRINTO (PRODUCT LAUNCH)

Sprinto is a security compliance platform that works with any cloud setup and helps monitor entity-level risks and controls from a single dashboard. It focuses on continuous compliance and healthy operational practices.

sprinto.com

APPLICATION INSPECTOR

Application Inspector is a source code analyzer built for surfacing features of interest and other characteristics to answer the question 'What's in the code?' quickly using static analysis with a JSON-based rules engine. It is ideal for scanning components before use or detecting feature-level changes.

github.comGitHub repository

GRAPHQL COP

GraphQL Cop is a Python utility that runs common security tests against GraphQL APIs. It can be incorporated into CI/CD and provides cURL requests for any findings.

github.comGitHub repository

04MISCELLANEOUS

R LANGUAGE FLAW ALLOWS CODE EXECUTION VIA RDS/RDX FILES

A critical vulnerability has been discovered in the R programming language, widely used for data analysis and AI/ML, that allows arbitrary code execution when deserializing specially crafted RDS and RDX files. The vulnerability exploits R's handling of serialization/deserialization, enabling attackers to embed malicious code in file metadata for execution upon opening the files.

bleepingcomputer.com4 min read

MILLIONS OF MALICIOUS 'IMAGELESS' CONTAINERS PLANTED ON DOCKER HUB OVER 5 YEARS

Cybersecurity researchers have uncovered multiple malicious campaigns targeting Docker Hub that planted millions of "imageless" containers over the past five years. Over 4 million repositories have been found to be empty except for documentation containing links to phishing or malware sites.

thehackernews.com3 min read