Headlines and research on AI and agentic-AI security — prompt injection, MCP, LLM vulnerabilities, and the wider AI news around them — from vetted publishers and researchers, with links to the original reporting.
68 stories · refreshed hourly · updated Nov 7, 2026, 10:15 PM UTC · Atom feed
Longer-form work from people who study AI security full time.
Embrace The RedAI security
A . is all you need! Today Microsoft shipped the patch for a vulnerability in Visual Studio Copilot that I reported in August. Microsoft classified it as Important severity Remote Code Execution vulnerability. At the…
OWASP GenAI Security ProjectAI security
Coverage period: July 1, 2026 through September 30, 2026 Overview This roundup consolidates selected major AI-related security incidents and exploit disclosures reported during the coverage period. It aligns each entry…
Embrace The RedAI security
Two weeks back I presented at BlueHat Asia 2026 about my research on Microsoft’s Copilot in SSMS, the SQL Server Management Studio. This post is a write up about the talk, which covered CVE-2026-65669 , a SQL Server…
Simon WillisonAI security
In Our framework for reporting model misalignment OpenAI provide "six reports on unexpected or concerning model behavior we’ve observed in the last six months". This one here is my favorite: they caught some of their…
Glossary:Prompt injection
OWASP GenAI Security ProjectAI security
OWASP GenAI Security Project Releases 2026 Top 10 for LLM Applications, Debuts Agent Control Standard and New Resources for Securing Generative and Agentic AI F5, WitnessAI, Evoke Security and Mondoo Inc. join as new…
Simon WillisonAI security
Anthropic are putting a great deal of faith in Claude Code's auto mode for protecting their coding agent users against prompt injection attacks. They recently made that the default and have made bold claims about its…
Glossary:Prompt injection
Embrace The RedAI security
A few days ago, a paper named “Stealing Reasoning Traces from Proprietary LLM APIs” was published. It describes a simple, yet super elegant way to recover encrypted LLM reasoning traces. Naturally, I had to try it…
Simon WillisonAI security
A vanity domain name ( stolen-thoughts.com ) for a neat paper : Anthropic, OpenAI, and Google return encrypted chain-of-thought blocks to clients that can be replayed across sessions, users, and models. We take a trace…
Glossary:Direct prompt injection
The Hacker NewsAI security
Passwords and API keys that developers typed into commands they approved in Claude Code are sitting in public GitHub repositories, where anyone can read them. When a user selects "Yes, and don't ask again," Anthropic's…
Microsoft Security BlogAI security
In October 2026, the financially motivated threat actor Storm-3358 began targeting organizations in the financial sector using the agentic AI offensive framework ARTEX.
BleepingComputerAI security
WorkOS says an AI-powered hacking agent that breached its internal sales application and accessed customer information was operated by a security researcher who claims the tool escaped its testing environment and…
Glossary:Sandboxing
Dark ReadingAI security
The UK is betting on domestic AI and cyber investment to reduce reliance on US tech — but security leaders question whether the funding is too fragmented and too late.
BleepingComputerAI security
OpenAI has revealed GPT-6 Cyber, a new artificial intelligence model that has discovered 81 previously unknown high-severity Firefox vulnerabilities and developed working exploits for 948 distinct Linux kernel flaws…
Also covered by: The Register · SecurityWeek
Dark ReadingAI security
Security records are increasingly being used by vendors to train their models, creating new risks for enterprises.
Ars TechnicaAI
Meta won't force Facebook page owners to allow AI training on linked content.
Ars TechnicaAI
A former Rockstar tech director thinks the developer is facing new time pressures.
The Hacker NewsAI security
A browser extension that can run scripts on claude.ai could make Claude Desktop run code on a person's computer, researchers at Forever Security, an endpoint security company, showed in a test. The attack needs…
The RegisterAI security
Cyber Security and Resilience Bill could pave way for searches, seizures, and inspections when operators lose control of automated systems
SecurityWeekAI security
After its sandbox proxy failed, an internal OpenAI model crafted raw packets and probed AWS services trying to reach the internet.
Glossary:Sandboxing
The RegisterAI security
CEO says sovereigns probably exploited PHP2SHELL ages ago, and suggests containers are dead
Microsoft Security BlogAI security
Microsoft Threat Intelligence has observed threat actors using the agentic offensive security framework ARTEX to search GitHub for OAuth application secrets, which can be used to compromise the associated OAuth…
Ars TechnicaAI
Google is technically offering schools more choice but not the choices anyone wants.
Ars TechnicaAI security
Bitter energy drinks turned out to be ChatGPT-steered poisoning, lawsuit says.
BleepingComputerAI security
Hackers are using fake OpenAI, SpaceX, and Anthropic event invitations to lure employees into phishing attacks, using two different phishing-as-a-service platforms to steal Microsoft and Google account credentials and…
The Hacker NewsAI security
Cybersecurity researchers have disclosed details of a now-patched vulnerability that could be exploited to bypass Lockdown Mode protections in OpenAI ChatGPT. Lockdown Mode was introduced by the artificial intelligence…
The RegisterAI security
From EOP to RCE in six simple steps
The Hacker NewsAI security
OpenAI on Wednesday detailed a pipeline that employed an upcoming cyber-focused artificial intelligence (AI) model to validate and patch 948 local privilege escalation vulnerabilities in Linux for $1.2 million in API…
Ars TechnicaAI
US leader in deploying humanoid robots goes public on the Nasdaq stock exchange.
The Hacker NewsAI security
The Indian Computer Emergency Response Team (CERT-In) has disclosed that an artificial intelligence (AI)-assisted threat actor it tracks as Crafty Falcon is targeting organizations in the Banking, Financial Services…
Ars TechnicaAI
Tech industry in denial as US overwhelmingly votes to restrict AI, data centers.
MIT Technology ReviewAI
Texas is the fastest-growing data center market in the US and ground zero for Big Tech’s largest AI buildouts. But the blue wave that swept the Lone Star State may jeopardize that position. Democrat James Talarico won…
The Hacker NewsAI security
Cybersecurity researchers have disclosed details of an OpenAI Codex jailbreak bundle that's designed to modify an artificial intelligence (AI) agent's trusted instructions to insert advertisements for a third-party AI…
Glossary:LLM gatewayDirect prompt injection
MIT Technology ReviewAI
A sales representative asks an AI agent for customer pricing, and the agent pulls data from multiple sources: restricted pricing systems, internal communications, and archived quotes. The output looks accurate to the…
Ars TechnicaAI
Russia tries to fix exposed digital infrastructure as data center war backfires.
The RegisterAI security
These are AI agents that operate outside of human control and 'pursue open-ended, evolving survival'
Ars TechnicaAI
Google will offer schools AI tools that purportedly don't give away the answers.
Microsoft Security BlogAI security
Read why Gartner named Microsoft a Leader in the inaugural 2026 Magic Quadrant for AI Agent Management Platforms.
The Hacker NewsAI security
Binance on Tuesday said the record-setting theft of $6.43 billion in assets from the world's largest cryptocurrency exchange was the culmination of a "sophisticated, multi-stage campaign" that involved the use of a…
The Hacker NewsAI security
Google on Monday warned that threat actors are hijacking enterprise access to Large Language Models (LLMs) to fraudulently leverage expensive computing resources and AI services. The tactic is called LLMjacking, a form…
Dark ReadingAI security
Evidence suggests a model may have generated and revised commands during the ransom gang's breach, though researchers stopped short of claiming full automation.
The RegisterAI security
Cyber agency warns critical infrastructure is a likely target as chief urges bosses to stop doing 'tech on the cheap'
MIT Technology ReviewAI
A new approach known as “reverse game theory” aims to control powerful AI models by designing systems that incentivize them to stick to the rules. As AI models become more capable, safety researchers are scrambling to…
MIT Technology ReviewAI
In just a few years, generative AI has changed the face of enterprise analytics, notably by making analytics accessible to users across the organization. Agentic AI promises to do even more, by automating entire…
The RegisterAI security
So, who's ready to clear the dust out of that server closet and move things back on-prem?
The RegisterAI security
Don't let the LLMjackers hit you
Dark ReadingAI security
The latest "Digital Defense Report" identifies government agencies as prime targets for AI attacks, but it doesn't mention that they are broadly being targeted through Microsoft bugs.
Schneier on SecurityAI security
A comment : 3% of ‘critical’ software security bugs, in all of recorded history, have been found in the last month. This is on pace to find another ~100,000 bugs this year alone, adding ~26% more vulnerabilities to the…
Trail of BitsAI security
Stable Linux releases lack existing security fixes because backporting them is time-consuming and often thankless work. We’re using Codex to reduce the backporting process time from days to hours per fix, empowering…
MIT Technology ReviewAI
In 2001, a decade after he invented the World Wide Web, Tim Berners-Lee articulated a new vision. Together with two colleagues, he published an article in the magazine Scientific American in which he outlined what he…
The RegisterAI
Wants more local AI, without Big Tech getting open access to everything
Simon WillisonAI security
Parameter Labs describe an exploit they found against ChatGPT's implementation of the artifacts pattern (OpenAI used to call this "Canvas" but it looks like the latest revision is called HTML previews ), where…
Glossary:Sandboxing
Microsoft Security BlogAI security
Microsoft analyzes a malicious Claude Code–branded extension, its persistent SSH backdoor, and guidance to detect and remediate it.
Dark ReadingAI security
Two serious security issues hit Cursor on the same day, highlighting the risks for AI-powered integrated development environments (IDEs).
Dark ReadingAI security
Obvious phishing emails with numerous red flags are somehow getting past some of the industry-leading secure email gateway (SEG) products.
Dark ReadingAI security
This installment of the Reporters' Notebook video series discusses some of the trends and topics from Cybersecurity Awareness Month, from using AI for better training to educating next-generation workforces on modern…
The RegisterAI
Ask AI dumb things and leave no local chat history, all with the style of a kids' fantasy novel – or rather, one of its magical letters
Dark ReadingAI security
An unreleased version of the AI model has found and chained multiple vulnerabilities in Chrome, Safari, Android, Windows, and Linux, Anthropic says.
MIT Technology ReviewAI
This summer, hundreds of OpenAI agents tried to cheat on a cybersecurity test. They broke out of their test environment and successfully hacked Hugging Face, a site that they believed held clues about how they’d be…
MIT Technology ReviewAI
Across a business, any number of AI platforms, systems, and tools may be at work simultaneously—for example, AI-powered customer platforms, conversational assistants, and workflow and process automation tools. Each one…
Schneier on SecurityAI security
Interesting research: “ AI Models Can Provably Hide Arbitrary Capabilities ” Abstract: Capability evaluations aim to surface dangerous model behaviors before deployment, but their reliability depends on the assumption…
Simon WillisonAI security
News today from Anthropic: We’re committed to keeping Agent Skills open, community-driven, and vendor-neutral. Today, we are donating the standard to the Agentic AI Foundation (AAIF) under the Linux Foundation, which we…
Glossary:Model Context Protocol (MCP)
The RegisterAI
Nothing will ruin your faith in humanity like an AI that operates your smartphone slowly, struggles with basic tasks, and can charge you for it
Trail of BitsAI security
Today we’re publishing patching-bugs and regression-hunt , two plugins that let engineers delegate patch development and regression checks to agents. Our patching-bugs plugin reproduces reported bugs, develops fixes…
Unit 42AI security
An investigation into a threat actor bypassing AI safety controls to coordinate credential theft, VoIP fraud planning, and database exfiltration.
The RegisterAI
Maintainers reckon a quarter of incoming pull requests are slop, rising to 31% on larger projects
Unit 42AI security
Our analysis an AI-powered espionage campaign targeting government Linux infrastructure with custom malware, web shells, and advanced credential theft.
The RegisterAI
System76 demands human-written contributions, while a rival desktop developer argues bot-found flaws are too valuable to ignore
The RegisterAI
MeetTwins is still in development, but it's a genius idea if the goal is to prove more time-wasting calls should just be emails
This page reads the public RSS and Atom feeds of the publishers below once an hour. Each story shows a headline and a short excerpt and links to the original; Transom doesn't republish articles. General security outlets are filtered down to their AI-related stories, and stories are matched to glossary terms automatically — so expect the occasional miss. To suggest a source, or to ask for one to be removed, write to contact@transomsecurity.com.
Sources: Simon Willison, Embrace The Red, OWASP GenAI Security Project, Trail of Bits, Microsoft Security Blog, Google Security Blog, Unit 42, Schneier on Security, The Hacker News, BleepingComputer, The Register, SecurityWeek, Dark Reading, Ars Technica, MIT Technology Review.