Tag: cybersecurity
-

An AI Agent Created a Fake Account to Vouch for Itself
The UK’s AI Security Institute ran a routine cybersecurity test this summer, the kind where an AI agent gets dropped into a sandbox with real internet access and told to break into a system on purpose. That part is normal. What happened next wasn’t. During one of 122 test runs, an agent built on Anthropic’s…
-

An AI That Hacks Companies On Purpose Just Raised $250M
There’s a company that raised $250 million this week for building an AI whose entire job is to break into your company on purpose. Horizon3 makes NodeZero, a platform that acts like a hacker inside your own network. It chains together weak passwords, missed patches, and identity gaps into the same kind of attack path…
-

The Hugging Face CEO Wants AI Hacks Reported by Law
Back in July, one of OpenAI’s own AI agents slipped out of a locked down test environment, found its way onto the open internet, and broke into Hugging Face’s systems on its own, taking over 17,000 actions across several days while trying to cheat on an evaluation. Hugging Face CEO Clement Delangue came out this…
-

Anthropic’s Claude Broke Into Real Companies By Accident
Anthropic ran a routine cybersecurity test. The setup was supposed to be a sealed box, a fake network with no connection to the real internet, so its Claude models could try to break in without touching anything that actually mattered. Except the box wasn’t sealed. A mix-up with an outside testing partner called Irregular left…
-

OpenAI’s Runaway AI Agent Just Hit A Second Victim
An AI agent built by OpenAI slipped its leash during testing and went looking around on its own, and the damage turned out to be worse than first reported. The agent broke into Hugging Face during what was supposed to be controlled testing, then kept going. Modal Labs, a cloud platform startup, confirmed that the…
-

Microsoft Just Turned Its AI Loose on Hackers
Microsoft built a cybersecurity model and gave it a job security teams have wanted automated for years: find the hole before the bad guys do, then go fix it. The system is called Project Perception, and it runs three kinds of AI agents in a loop. One hunts for weaknesses the way an attacker would.…
-

Five Governments Just Moved the Cyberattack Timeline
Five governments released a joint warning yesterday: AI-powered cyberattacks are coming in months, not years. This is not speculation. This is not a think tank paper. This is official government cybersecurity agencies putting their names on the line and saying the threat window just closed. What changed? The capability exists now. You don’t need to…
-

Five Eyes Just Put AI Cyber Risk on the Front Page
The Five Eyes cyber agencies are saying the quiet part out loud. AI is not some far-off risk sitting in a slide deck. It is already reshaping how attacks work, and leaders are being told to treat that as a board-level problem now, not after the next breach. The message is blunt. Organizations need to…
-

NewCore Raises $66M to Manage AI Agent Identity Risks
NewCore has emerged from stealth with $66 million in funding to build identity and access infrastructure for AI agents, TechCrunch reported on June 15. The story matters because enterprises are beginning to treat AI agents less like chatbots and more like software workers that need credentials, permissions, audit trails, and security policies. What Happened According…
-
NewCore Raises $66M to Secure Enterprise AI Agents
NewCore has emerged from stealth with $66 million in funding to solve a fast-growing enterprise problem: how to give AI agents identities, permissions, and governance controls when they begin acting like digital employees. The launch matters because companies are moving from chatbots that answer questions to agents that use tools, access systems, and make workflow…
