An unknown threat actor has been observed using a large language model (LLM) agent to conduct post-compromise actions after obtaining initial access following the exploitation of a publicly-accessible Marimo network using a recently disclosed vulnerability. “The attacker compromised an internet-reachable Marimo notebook via CVE-2026-39987, extracted two cloud credentials from the compromised
Tag: language
Global Security News, AI
Frontier AI models collapse under multi-turn AI attacks, Cisco finds
Attackers who probe large language models rarely give up after one refusal. They reframe, build context across turns, adopt personas, and escalate gradually. New research from Cisco’s AI threat intelligence team finds that the safety benchmarks used across the industry miss almost all of this behavior, and the gap between published scores and observed resilience…
AI, Global Security News, Network Security
When your AI assistant has the keys to production
Large language models in operational roles query telemetry, propose configuration changes, and in some deployments execute those changes against live infrastructure. Ticket drafting and alert summarization were the starting point. Vendors describe this work as autonomous remediation or self-healing infrastructure. A recent survey on agentic AI in network and IT operations gives it a more…
AI, Cloud Security, Cybersecurity, Data Breaches, Europe, Global Security News, Government & Policy, malware, Network Security, Risk Management
Daybreak is OpenAI’s answer to the AI arms race in cybersecurity
OpenAI has unveiled Daybreak, a cybersecurity initiative that combines the company’s large language models with its Codex agentic framework to help organizations identify, patch, and validate software vulnerabilities across the development lifecycle. The platform is built around three model tiers: GPT-5.5 for general-purpose use, GPT-5.5 with Trusted Access for Cyber for verified defensive security workflows,…
AI, Compliance, Global Security News
AI is ready to take over Python programming, but not much else
Tests of how well 19 large language models (LLMs) complete and perform complicated multi-step tasks has shown that they are both error-prone and, in many cases, unreliable. The findings are contained a preprint paper, LLMs Corrupt Your Documents When You Delegate, written by Microsoft researchers Philippe Laban, Tobias Schnabel and Jennifer Neville based on a…
AI, Global Security News
RubyGems Suspends New Signups After Hundreds of Malicious Packages Are Uploaded
RubyGems, the standard package manager for the Ruby programming language, has temporarily paused account sign ups following what has been described as a “major malicious attack.” “We’re dealing with a major malicious attack on Ruby Gems right now,” Maciej Mensfeld, senior product manager for software supply chain security at Mend.io, said in a post on…
AI, Exploits, Global Security News
Hackers Use AI for Exploit Development, Attack Automation
Cyber adversaries have long used AI, but now attackers are using large language models to develop exploits and orchestrate complex attacks.
AI, Compliance, Cybersecurity, Exploits, Global Security News, Network Security, Risk Management
Cisco Reveals Security Gaps in Vision Language Models
Vision language models (VLMs) continue to expand the capabilities of artificial intelligence by combining image and text understanding into a single system. However, recent research from Cisco into typographic prompt injection attacks highlights significant weaknesses in how these models interpret and secure visual information. The second installment of Reading Between the Pixels explores how small…
AI, Global Security News
Writers Are Going to Extremes to Prove They Didn’t Use AI
People are adding typos, aggressively casual language and references to ‘The Office’ to stay ahead of armchair detectors.
AI, Apps, Global Security News, Network Security, privacy
Open-source privacy proxy masks PII before prompts reach external AI services
Enterprise developers routinely send prompts to external large language models that contain customer emails, support transcripts, and other identifying information, often without a sanitization layer between the application and the API. Dataiku has released Kiji Privacy Proxy, an open-source local gateway that detects and masks personally identifiable information before requests leave the network. The tool…
AI, Global Security News
Automated LLM red teaming gets a learning layer
Automated red teaming of large language models has settled into a familiar pattern over the past two years. An attacker model generates jailbreak attempts against a target model, an evaluator scores the results, and the cycle repeats. Two approaches dominate. One asks the attacker to invent strategies through trial and error, which tends to produce…
AI, Global Security News
Google Clears Pentagon to Use AI Tools in Classified Settings
Tech giant added language to contract to say its AI wasn’t intended for domestic mass surveillance or fully autonomous weapons.
AI, Global Security News
Google Favors General‑Purpose Gemini Models Over Cybersecurity‑Specific AI
Google Cloud’s COO advocated for combining general-purpose frontier large language models with task-specific AI agents
AI, Global Security News
PCI SSC Launches Enhanced Language Microsites for Global Audience
The PCI Security Standards Council (PCI SSC) has announced the launch of newly redesigned language microsites, delivering a more accessible, structured, and user-friendly experience for global stakeholders. These updates mark a significant step forward in making PCI resources more readily available to non-English-speaking audiences by providing a consistent user experience across languages.
AI, Global Security News
Affinda Launches No-code AI Integration Agent, Unlocking Document Automation At Scale
Agent delivers seamless integration in minutes using natural language – enabling high-stakes organisations of all sizes to move faster and tech teams to prove concepts quickly
AI, Cybersecurity, Global Security News, Government & Policy, Network Security
Testing reveals Claude Mythos’s offensive capabilities and limits
Could Claude Mythos Preview, Anthropic’s latest large language model, be leveraged for fully automated cyber attacks? The UK government’s AI Security Institute (AISI) tested its capability to successfully engage in capture-the-flag (CTF) challenges and multi-step attack scenarios, and found that that while its cybersecurity capabilities exceed those of previously available models, it can’t reliably execute…
AI, Global Security News
Google study finds LLMs are embedded at every stage of abuse detection
Online platforms are running large language models at every stage of LLM content moderation, from generating training data to auditing their own systems for bias. Researchers at Google mapped how this is happening across what the authors call the Abuse Detection Lifecycle, a four-stage framework covering labeling, detection, review and appeals, and auditing. Earlier moderation…
Global Security News
A nearly undetectable LLM attack needs only a handful of poisoned samples
Prompt engineering has become a standard part of how large language models are deployed in production, and it introduces an attack surface most organizations have not yet addressed. Researchers have developed and tested a prompt-based backdoor attack method, called ProAttack, that achieves attack success rates approaching 100% on multiple text classification benchmarks without altering sample…
AI, Global Security News
Rubrik Rolls Out Industry’s First Semantic AI Governance Engine
Domain-Specific Small Language Model Accelerates Trusted AI Agent Deployment and Control
AI, Global Security News
Google’s TurboQuant cuts AI memory use without losing accuracy
Large language models carry a persistent scaling problem. As context windows grow, the memory required to store key-value (KV) caches expands proportionally, consuming GPU memory and slowing inference. A team at Google Research has developed three compression algorithms: TurboQuant, PolarQuant, and Quantized Johnson-Lindenstrauss (QJL). All three are designed to compress those caches aggressively without degrading…
AI, Cybersecurity, Global Security News
Rethinking Song Creation Through Language Driven Systems
In this post, I will talk about rethinking song creation through language driven systems. The first time I experimented with an AI Music Generator, it felt less like using software and more like having a conversation. Instead of arranging tracks or selecting plugins, I was describing intentions—tone, pacing, emotion. That experience highlighted a persistent issue…
AI, Global Security News
Llamafile, Mozilla’s portable LLM runner, gets GPU support and a rebuilt core
Running a large language model on a single machine without cloud access or a container runtime remains a priority for practitioners working in air-gapped or resource-constrained environments. Llamafile, Mozilla-AI’s project for packaging and running LLMs as self-contained executables, has received its most significant architectural overhaul to date with version 0.10.0. A rebuild from the ground…
Apps, Endpoint, Global Security News, Risk Management
How Exposed Endpoints Increase Risk Across LLM Infrastructure
As more organizations run their own Large Language Models (LLMs), they are also deploying more internal services and Application Programming Interfaces (APIs) to support those models. Modern security risks are being introduced less from the models themselves and more from the infrastructure that serves, connects and automates the model. Each new LLM endpoint expands the…
AI, Apps, Compliance, Cybersecurity, Global Security News, Risk Management
AI FOMO: How Pressure to Adopt AI is Outpacing Understanding
AI – or large language models (LLMs) – is introducing new attack surfaces, despite the new capabilities that the technology promises. The new threats it is introducing, including prompt injection, deepfakes, and alignment risks, are huge security concerns at a strategic level. AI FOMO is driving enterprise adoption before risk mitigation At the Genetec Global…
AI, Europe, Global Security News
Cohere’s Tiny Aya Models Bring 70+ Languages to Offline AI
In the world of generative AI, language support has often been a luxury reserved for a handful of global languages. That’s changing fast. Cohere just unveiled a suite of open multilingual models designed to push AI out of data centers and into everyday devices while embracing linguistic diversity at scale. These “Tiny Aya” models underscore…
AI, Exploits, Global Security News, Politics
Why are AI leaders fleeing?
Normally, when big-name talent leaves Silicon Valley giants, the PR language is vanilla: they’re headed for a “new chapter” or “grateful for the journey” — or maybe there’s some vague hints about a stealth startup. In the world of AI, though, recent exits read more like a whistleblower warnings. Over the past couple of weeks,…
AI, Europe, Global Security News, Network Security
T-Mobile offers a reason to call, not text: simultaneous translation
Could language barriers be a thing of the past? Various companies are looking to bring AI to the party to solve this linguistic issue and maybe help bring voice once again to the fore. T-Mobile is preparing to offer network-based real-time translation services in more than 50 languages. The new service, currently in beta testing,…
AI, Global Security News
Azul survey shows significant increase in Java being used to code AI functionality
GUEST RESEARCH: More and more organisations are using Java as foundational language for AI development, with Azul’s 2026 State of Java Survey & Report revealing a significant increase.
AI, Artificial Intelligence, Generative AI, Global Security News, Risk Management
AI chatbots are worse than search engines for medical advice
There is a clear gap between the theoretical medical knowledge of large language models (LLMs) and their practical usefulness for patients, according not a new study from the Oxford Internet Institute and the Nuffield Department of Primary Care Health Sciences at the University of Oxford. The research, conducted in collaboration with MLCommons and other institutions,…
AI, APAC, Apps, Global Security News, Infrastructure, Risk Management
Java Adoption Accelerates for AI Workloads, Azul Survey Finds
Java is increasingly being positioned as a core language for enterprise AI development, even as organizations accelerate plans to move away from Oracle Java due to pricing and licensing concerns, according to Azul’s newly released 2026 State of Java Survey & Report. The annual study is based on responses from more than 2,000 Java professionals…
AI, Global Security News
Why AI Chatbots Can’t Be Trusted for Financial Advice: They’re Sociopaths
Today’s “large language models” like ChatGPT don’t have the training to act in users’ best interest, but an MIT professor hopes to teach them how.
AI, Global Security News
Why AI Chatbots Can’t Be Trusted for Financial Advice: They’re Sociopaths
Today’s “large language models” like ChatGPT don’t have the training to act in users’ best interest, but an MIT professor hopes to teach them how.
AI, Apps, Artificial Intelligence, Compliance, Cybersecurity, Data Breaches, Exploits, Global Security News, Network Security, News, Risk Management, Threats
Claude Opus 4.6 Exposes Hundreds of Open-Source Vulnerabilities
Artificial intelligence firm Anthropic says its newest large language model, Claude Opus 4.6, has identified more than 500 previously unknown high-severity vulnerabilities across widely used open-source libraries. It “… reads and reasons about code the way a human researcher would — looking at past fixes to find similar bugs that weren’t addressed, spotting patterns that…
AI, Apps, Artificial Intelligence, Compliance, Cybersecurity, Data Breaches, Exploits, Global Security News, Network Security, News, Risk Management, Threats
Claude Opus 4.6 Exposes Hundreds of Open-Source Vulnerabilities
Artificial intelligence firm Anthropic says its newest large language model, Claude Opus 4.6, has identified more than 500 previously unknown high-severity vulnerabilities across widely used open-source libraries. It “… reads and reasons about code the way a human researcher would — looking at past fixes to find similar bugs that weren’t addressed, spotting patterns that…
AI, Apps, Artificial Intelligence, Compliance, Cybersecurity, Data Breaches, Exploits, Global Security News, Network Security, News, Risk Management, Threats
Claude Opus 4.6 Exposes Hundreds of Open-Source Vulnerabilities
Artificial intelligence firm Anthropic says its newest large language model, Claude Opus 4.6, has identified more than 500 previously unknown high-severity vulnerabilities across widely used open-source libraries. It “… reads and reasons about code the way a human researcher would — looking at past fixes to find similar bugs that weren’t addressed, spotting patterns that…
AI, Apps, Artificial Intelligence, Compliance, Cybersecurity, Data Breaches, Exploits, Global Security News, Network Security, News, Risk Management, Threats
Claude Opus 4.6 Exposes Hundreds of Open-Source Vulnerabilities
Artificial intelligence firm Anthropic says its newest large language model, Claude Opus 4.6, has identified more than 500 previously unknown high-severity vulnerabilities across widely used open-source libraries. It “… reads and reasons about code the way a human researcher would — looking at past fixes to find similar bugs that weren’t addressed, spotting patterns that…
AI, Apps, Artificial Intelligence, Compliance, Cybersecurity, Data Breaches, Exploits, Global Security News, Network Security, News, Risk Management, Threats
Claude Opus 4.6 Exposes Hundreds of Open-Source Vulnerabilities
Artificial intelligence firm Anthropic says its newest large language model, Claude Opus 4.6, has identified more than 500 previously unknown high-severity vulnerabilities across widely used open-source libraries. It “… reads and reasons about code the way a human researcher would — looking at past fixes to find similar bugs that weren’t addressed, spotting patterns that…
AI, Global Security News
Claude AI finds 500 high-severity software vulnerabilities
Anthropic only released its latest large language model, Claude Opus 4.6, on Thursday, but it has already been using it behind the scenes to identify zero-day vulnerabilities in open-source software. In the trial, it put Claude inside a virtual machine with access to the latest versions of open source projects, and provided it with a…
AI, Global Security News
Claude Opus 4.6 Finds 500+ High-Severity Flaws Across Major Open-Source Libraries
Artificial intelligence (AI) company Anthropic revealed that its latest large language model (LLM), Claude Opus 4.6, has found more than 500 previously unknown high-severity security flaws in open-source libraries, including Ghostscript, OpenSC, and CGIF. Claude Opus 4.6, which was launched on Thursday, comes with improved coding skills, including code review and debugging capabilities, along