Latest

Webinar warns of unowned AI agents as a new insider risk

A session promoted by The Hacker News on 22 October 2026 argues that autonomous agents holding credentials are running inside company networks with no named owner.

22 Oct 2026

Incidents

AI agents helped Chinese crew hit 2,250 Microsoft 365 tenants

Microsoft says Storm-0940 used an open source AI pentesting framework and stolen ConnectWise application credentials to reach privileged cloud access in under three hours.

22 Oct 2026

Vulnerabilities

OpenCode flaw let malicious MCP servers run code at sign-in

Anomaly fixed the AI coding agent in version 1.18.33 after researchers showed a hostile tool server could run commands on Windows and WSL machines.

22 Oct 2026

A puppet's strings slip through a cage's bars, allowing its wooden hands to reach forbidden doors while its master watches helplessly from behind locked gates.

Vulnerabilities

OpenClaw patches browser flaw that bypassed blocked URLs

A GitHub advisory published on 21 October 2026 says OpenClaw's browser debugging discovery could connect to network destinations the agent's own policy was meant to block.

21 Oct 2026

Incidents

Microsoft links ConnectWise supply-chain breach to AI attack tool

Microsoft says the China-state aligned actor Storm-0940 used the open-source agentic framework ARTEX to automate attacks on hundreds of organisations from September 2026.

21 Oct 2026

Incidents

Trojanised LiteLLM release steals cloud and AI provider keys

A malicious version 1.107.0 of the popular AI gateway reached PyPI and Docker registries on 21 October 2026, carrying an infostealer and a remote access trojan.

21 Oct 2026

A robot's hand reaches through an open door into neighboring rooms, grabbing files from shelves it was never meant to access, while guards stand motionless at the threshold.

Vulnerabilities

MCP tool in djust reads files outside the project directory

A GitHub advisory published on 21 October 2026 says a djust tool opens any absolute file path an agent asks for, with no confinement check.

21 Oct 2026

Vendors

Trail of Bits ships agent plugins to police AI-written patches

The security firm released two Codex plugins on 21 October 2026 after finding 21 regressions in more than 130 agent-assisted patches merged upstream.

21 Oct 2026

A hunter robot follows glowing false trails through a dark forest while real prey escapes untouched deeper in the woods.

Vulnerabilities

‘Context bombs’ send attacking AI agents chasing fake targets

Research reported by SecurityWeek on 21 October 2026 says planted decoy prompts cut the success rate of AI-driven attacks by up to 91% in testing.

21 Oct 2026

Incidents

Unit 42 says attacker used AI to help plan credential theft

Palo Alto Networks researchers published an investigation on 21 October 2026 into a threat actor who bypassed an AI system's safety controls across several criminal campaigns.

21 Oct 2026

Incidents

Malicious LiteLLM release ships credential stealer and backdoor

Version 1.107.0 of the popular Python package was uploaded to PyPI and Docker on 21 October 2026 and has since been quarantined.

21 Oct 2026

Explainers

Snyk says model risk is set by what an agent can reach

A post from the security vendor argues that a model's own risk rating means little without knowing the data, tools and actions its application can touch.

21 Oct 2026

Vendors

Asymptote Labs raises $5.5m for AI agent runtime security

The seed round, led by Tower Research Ventures, will fund commercial development of the startup's Beacon platform for watching agents while they run.

21 Oct 2026

Vulnerabilities

AI models can plan attacks on industrial systems, study finds

Frenos ran nearly 17,000 tests on ten AI models and found the best produced workable attack plans, but with inconsistent results and frequent gaps.

21 Oct 2026

Vendors

Nvidia pitches OpenShell blueprint to keep AI agents in check

The chipmaker's reference architecture ties agent sandboxing, credential brokering and behaviour monitoring to its own Vera, Rubin and BlueField hardware.

21 Oct 2026

Explainers

The harness, not the model, is where AI pentesting gets hard

A piece published on 21 October 2026 argues the engineering in agentic penetration testing sits in the controls wrapped around the model, not the model itself.

21 Oct 2026

A factory assembly line stamps identical circuit boards, but one stream diverges into a shadowy fork, producing components with invisible cracks only discoverable under strain.

Vulnerabilities

Chinese AI model writes weaker code for US defence tasks

Researchers reported on 21 October 2026 that the model produced code 19% more likely to be hackable when told the customer was a US defence contractor.

21 Oct 2026

A marionette puppet's strings feed directly into a massive machine's control lever, bypassing all safety gates between them.

Vulnerabilities

Prompt injection flaw lets attackers run root commands in Vibe-Trading

A vulnerability disclosed on 21 October 2026 affects HKUDS Vibe-Trading from version 0.1.0 before 0.1.7, with shell commands passed straight to the operating system.

21 Oct 2026

A digital key slips from one messenger's hand into another's as they pass through an unsecured doorway, the path splitting unexpectedly toward a waiting stranger.

Vulnerabilities

A2A Java SDK leaks agent API keys through HTTP redirects

Two advisories published on 20 October 2026 say a malicious agent can push a client's API key to a third-party server, and that an earlier fix missed the Android path.

20 Oct 2026

Incidents

Wiz details phishing campaign that hijacks cloud billing for AI use

Researchers say a single click on a Google Search advert can end with stolen credentials and hundreds of thousands of dollars in unauthorised AI API charges.

20 Oct 2026