CISA, Australia, and Partners Author Joint Guidance on Securely Integrating Artificial Intelligence in Operational Technology

CISA and the Australian Signals Directorate’s Australian Cyber Security Centre, in collaboration with federal and international partners, have released new cybersecurity guidance: Principles for the Secure Integration of Artificial Intelligence in Operational Technology. This guidance aims to hel...

The vulnerability of the LZ4_decompress_fast() function in the lz4-java data compression library allows a hacker to cause a service failure and disclose confidential information.

The vulnerability of the LZ4decompressfast function in the lz4-java data compression library is related to reading beyond the buffer boundaries in memory. Exploiting this vulnerability could allow a remote attacker to cause service interruptions and expose sensitive information...

RockyLinux 9 : kernel (RLSA-2025:20518)

The remote RockyLinux 9 host has packages installed that are affected by multiple vulnerabilities as referenced in the RLSA-2025:20518 advisory. kernel: can: isotp: fix potential CAN frame reception race in isotprcv CVE-2022-48830 kernel: soc: qcom: cmd-db: Map shared memory as WC, not WB...

Whispering poetry at AI can make it break its own rules

Most of the big AI makers don't like people using their models for unsavory activity. Ask one of the mainstream AI models how to make a bomb or create nerve gas and you'll get the standard "I don't help people do harmful things" response. That has spawned a cat-and-mouse game of people who try to...

Race Condition within a Thread

Overview mcp-mesh is a Kubernetes-native platform for distributed MCP applications Affected versions of this package are vulnerable to Race Condition within a Thread due to a Race Condition in Access Control. The agent shutdown process fails to atomically remove agents from the registry before...

The vulnerability in the module sound/soc/intel/boards/bytcr_rt5640.c of the Linux operating system allows a hacker to cause a service failure.

The vulnerability in the sound/soc/intel/boards/bytcrrt5640.c kernel of the Linux operating system relates to reading data outside the buffer boundaries in memory. Exploiting this vulnerability can allow an attacker to cause a system failure...

Learning the Wrong Lessons: Syntactic-Domain Spurious Correlations in Language Models

Whitepaper from researchers at MIT, Northeastern University, and Meta. For an LLM to correctly respond to an instruction it must understand both the semantics and the domain i.e., subject area of a given task-instruction pair. However, syntax can also convey implicit information Recent work shows...

A Wolf in Sheep's Clothing: Bypassing Commercial LLM Guardrails Via Harmless Prompt Weaving and Adaptive Tree Search

Large language models LLMs remain vulnerable to jailbreak attacks that bypass safety guardrails to elicit harmful outputs. Existing approaches overwhelmingly operate within the prompt-optimization paradigm: whether through traditional algorithmic search or recent agent-based workflows, the...

Malicious code in nova-angular-project (npm)

--- -= Per source details. Do not edit below this line.=- Source: amazon-inspector 6c1edbc661355ab22d5857395431ed8c7eb4bd73b77e1080a8cdb5ad1c386a8e The package nova-angular-project was found to contain malicious code...

kernel: mm/memory-failure: fix VM_BUG_ON_PAGE(PagePoisoned(page)) when unpoison memory

A use-after-free memory bug exists in the linux kernel, such that unpoisonmemory tries to check the PGHWPoison flags of an uninitialized page. So VMBUGONPAGEPagePoisonedpage is triggered, leading to damage to the system availability and integrity...

The WIRED Guide to Digital Opsec for Teens

Practicing good “operations security” is essential to staying safe online. Here's a complete guide for teenagers and anyone else who wants to button up their digital lives...

Red Teaming Large Reasoning Models

Large Reasoning Models LRMs have emerged as a powerful advancement in multi-step reasoning tasks, offering enhanced transparency and logical consistency through explicit chains of thought CoT. However, these models introduce novel safety and reliability risks, such as CoT-hijacking and...

LZ4 Java Compression has Out-of-bounds memory operations which can cause DoS

Out-of-bounds memory operations in org.lz4:lz4-java 1.8.0 and earlier allow remote attackers to cause denial of service and read adjacent memory via untrusted compressed input. This is fixed in a forked release: at.yawk.lz4:lz4-java version 1.8.1. The original project has been archived:...

EUVD-2025-199880

Out-of-bounds memory operations in org.lz4:lz4-java 1.8.0 and earlier allow remote attackers to cause denial of service and read adjacent memory via untrusted compressed input...

Prompt Injection Through Poetry

In a new paper, "Adversarial Poetry as a Universal Single-Turn Jailbreak Mechanism in Large Language Models," researchers found that turning LLM prompts into poetry resulted in jailbreaking the models: Abstract : We present evidence that adversarial poetry functions as a universal single-turn...

`Reader::open_mmap` unsoundly marks unsafe memmap operation as safe

maxminddb prior to version 0.27 declared Reader::openmmap as safe despite wrapping an inherently unsafe memmap2 operation with no extra step done to guarantee safety. This could have led to undefined behaviour if the file were to be modified on disk while the memory map was still active...

PT-2025-52888

Name of the Vulnerable Software and Affected Versions Linux kernel affected versions not specified Description An issue exists in the Linux kernel related to out-of-bounds memory access within the ch341 transfer one function of the SPI CH341 driver. The len variable calculation includes a command...

Webinar: Learn to Spot Risks and Patch Safely with Community-Maintained Tools

If you're using community tools like Chocolatey or Winget to keep systems updated, you're not alone. These platforms are fast, flexible, and easy to work with—making them favorites for IT teams. But there's a catch... The very tools that make your job easier might also be the reason your systems...

AI Meeting Assistants Are Rising – But Is Your Data Safe? A Deep Look at TicNote AI

AI meeting assistants have become essential tools for professionals who want fast, accurate, and automated transcription. Yet behind…...

CVE-2025-13601

A heap-based buffer overflow problem was found in glib through an incorrect calculation of buffer size in the gescapeuristring function. If the string to escape contains a very large number of unacceptable characters which would need escaping, the calculation of the length of the escaped string...