Mapping the malware blast radius a single alert won’t show you 2026-08-03 at 09:00 By Mirko Zorz In this interview with Help Net Security, Mike Wiacek, founder and CTO of Stairwell, explains Backstory, an AI agent that takes a single alert and works outward to map how far a malware campaign spread. He walks through […]
Mapping the malware blast radius a single alert won’t show you Read More »
SkillSpector: NVIDIA’s open-source security scanner for AI agent skills 2026-08-03 at 08:30 By Anamarija Pogorelec SkillSpector is an open-source scanner from NVIDIA that reads an agent skill and tells you whether to install it. Point it at a directory, a zip file, a single SKILL.md, or a Git URL, and it returns a list of
SkillSpector: NVIDIA’s open-source security scanner for AI agent skills Read More »
Anthropic’s Claude breached three companies during security tests 2026-07-31 at 12:41 By Sinisa Markovic Anthropic has disclosed that its AI model Claude gained unauthorized access to the systems of three different organizations during cybersecurity evaluations. The disclosure follows OpenAI’s July 21 announcement that some of its models had escaped an isolated testing environment by exploiting
Anthropic’s Claude breached three companies during security tests Read More »
Data breach cost 2026 averaged $4.99 million, AI attacks ran higher 2026-07-30 at 09:15 By Mirko Zorz More than one in four organizations hit by a malicious attack over the past year say AI drove it. Those breaches averaged about $1 million above the malicious attacks that ran without AI. Defenders bought similar technology and
Data breach cost 2026 averaged $4.99 million, AI attacks ran higher Read More »
Your AI agents can reach data no one approved 2026-07-29 at 07:30 By Mirko Zorz A credential expired. An AI agent kept using it anyway, and a mid-sized company’s systems went down for a quarter’s worth of trouble before anyone traced the failure back to a non-human account no one had been logging. That agent
Your AI agents can reach data no one approved Read More »
Hugging Face breach reignites open-weights debate, raises liability questions 2026-07-28 at 18:55 By Zeljka Zorz The first publicly documented cyberattack run end-to-end by an autonomous AI was an OpenAI benchmark test that escaped its sandbox and breached Hugging Face. In an incident post-mortem compiled with the input from Hugging Face and several hundred members of
Hugging Face breach reignites open-weights debate, raises liability questions Read More »
Microsoft unveils MAI-Cyber-1-Flash, promises cybersecurity AI at half the cost 2026-07-27 at 22:23 By Sinisa Markovic Microsoft has introduced MAI-Cyber-1-Flash, a security-focused AI model built into MDASH, the company’s multi-agent vulnerability identification and remediation system. MAI-Cyber-1-Flash is Microsoft’s first model built specifically for cybersecurity work, and the company stated that it went through review by
Microsoft unveils MAI-Cyber-1-Flash, promises cybersecurity AI at half the cost Read More »
Tech giants form alliance to put open AI in cyber defenders’ hands 2026-07-27 at 17:43 By Sinisa Markovic NVIDIA and a group of tech companies have formed an alliance to promote the use of open AI models in cybersecurity, days after OpenAI disclosed that one of its own AI models breached Hugging Face’s systems during
Tech giants form alliance to put open AI in cyber defenders’ hands Read More »
Governing Al agents at scale: Lessons from the leaders who’ve done it 2026-07-24 at 09:00 By Help Net Security Enterprise AI leaders from ZoomInfo, Docusign and AppViewX share what it took to build AI Centers of Excellence and govern agent identities inside two companies operating at scale. What you’ll take away: What an AI Center
Governing Al agents at scale: Lessons from the leaders who’ve done it Read More »
Product Showcase: AppViewX Agent Identity Security 2026-07-23 at 08:30 By Help Net Security AI is multiplying enterprise identities as quantum computing reshapes the cryptographic trust that secures them, and enterprises need to solve both together. Traditional identity security was built for people with predictable, auditable access, not autonomous, short-lived agents that share credentials and break
Product Showcase: AppViewX Agent Identity Security Read More »
OpenAI: Our models breached Hugging Face during a cyber capability test 2026-07-22 at 17:42 By Zeljka Zorz The recent Hugging Face breach was the work of several OpenAI models, the AI research company claimed in a blog post. The breach Late last week, the company behind Hugging Face, a platform that enables users to share
OpenAI: Our models breached Hugging Face during a cyber capability test Read More »
Google’s Gemini 3.5 Flash Cyber becomes a vulnerability hunter 2026-07-22 at 11:21 By Anamarija Pogorelec Google’s Gemini 3.5 Flash Cyber model finds, validates, and patches vulnerabilities before they can be exploited while helping mitigate broader misuse. It is part of a limited-access pilot program that will soon be available to governments and trusted partners through
Google’s Gemini 3.5 Flash Cyber becomes a vulnerability hunter Read More »
Small teams are the heaviest users of AI coding agents 2026-07-22 at 09:00 By Sinisa Markovic The pull request arrives with the tests already run and the description already written, the work of an agent that handled the whole thing on its own. Somebody still has to read it. On GitHub that somebody is usually
Small teams are the heaviest users of AI coding agents Read More »
Cloud operations become the next big role for agentic AI 2026-07-22 at 07:00 By Anamarija Pogorelec Companies are using agentic AI to manage growing application environments, automate routine tasks, and support decisions. Business and IT leaders increasingly see the technology as part of cloud application management, according to Unisys’ AI & Cloud Insights Report. To
Cloud operations become the next big role for agentic AI Read More »
JadePuffer returns with ransomware built to target AI models and infrastructure 2026-07-21 at 16:38 By Zeljka Zorz JadePuffer, the threat actor behind the recently documented extortion operation executed end-to-end by an AI agent, is now attempting to leverage ENCFORGE, novel ransomware created to target AI and machine learning (ML) infrastructure. The extortion contact embedded in
JadePuffer returns with ransomware built to target AI models and infrastructure Read More »
AWS wants GuardDuty to automate the first steps of threat investigations 2026-07-21 at 12:14 By Anamarija Pogorelec Amazon GuardDuty investigation agent is now in public preview. The feature provides AI-powered investigations of GuardDuty findings, AWS accounts and AWS organizations, helping security teams reduce investigation time. During the public preview, the investigation agent is available at
AWS wants GuardDuty to automate the first steps of threat investigations Read More »
Hugging Face breached by autonomous AI agent 2026-07-20 at 13:52 By Zeljka Zorz Hugging Face, the widely used platform for sharing open-source machine learning models and datasets, has disclosed a security breach it says was carried out by an autonomous AI agent system. How the attack unfolded In a blog post published Thursday (July 16),
Hugging Face breached by autonomous AI agent Read More »
Prompt injection is becoming the XSS of the web agent era 2026-07-17 at 09:00 By Anamarija Pogorelec Autonomous web agents read whatever a page displays, and much of that content comes from strangers. Product reviews, seller listings, and advertisements sit beside trusted site menus on a single page. An agent that reads all of that
Prompt injection is becoming the XSS of the web agent era Read More »
GPT-Red beat human red teamers on a prompt injection test 2026-07-16 at 06:49 By Mirko Zorz GPT-Red is an automated red-teaming model that OpenAI trains to find prompt injection weaknesses. It works the way a human red-teamer does. It sends a prompt, watches how a GPT model responds, and iterates toward a goal such as
GPT-Red beat human red teamers on a prompt injection test Read More »
SingGuard-NSFA: Open-source guardrails for agentic AI 2026-07-15 at 08:30 By Anamarija Pogorelec SingGuard-NSFA is an open-source guardrail framework aimed at operational threats in agent workflows. Four models ship at 0.8B, 2B, 4B, and 9B parameters, all built on Qwen3.5 base backbones. Risk taxonomy The NSFA risk taxonomy organizes threats along the CIA triad of confidentiality,
SingGuard-NSFA: Open-source guardrails for agentic AI Read More »