Tuesday 17 November 2026 · UTC
Global

Your region, kept on this device.

The week in security

Patch, watch and understand: up to eight stories from the past seven days.

Latest included publication: .

Automatically selected from published coverage, prioritising known exploitation, critical severity and source coverage. Selection is not a human review. Each story links to its evidence and available actions.

Daily brief · Weekly brief · Get the Friday email · Check your exposure

Patch

Check affected versions and vendor instructions before making changes.

Cyber Security News· · Critical severity

Google patches actively exploited Chrome zero-day CVE-2026-125350 in November update

issued a Stable update on 9 November fixing nine security issues, including CVE-2026-125350, a high-severity use-after-free bug in Chrome's XML component that Google says is exploited in the wild. The fixed builds are 157.0.8103.50/.51 for Windows and Mac and 157.0.8103.50 for Linux, with Android builds at 157.0.8103.50 on Google Play. HKCERT rates the overall risk as Extremely High, citing possible denial of service, security restriction bypass and remote code execution.

Watch

Incidents and campaigns to assess against your own exposure.

Cyber Security News·

SCKit worm abuses ScreenConnect to install unauthorised remote access and target technicians

ConnectWise reported an active campaign from 11 November 2026 in which attackers gained unauthorised access to cloud-hosted ScreenConnect instances used by managed service providers, and Threat Intelligence confirmed this activity. The toolkit installs persistence through scheduled tasks and Run registry entries, deploys to deliver a payload to connecting technicians, and uses hijacked technician sessions to push payloads to other managed devices. Removing the unauthorised client alone does not remove the persistence or invalidate exposed web sessions.

Cyber Security News·

GhostFrame malware targets VMware ESXi servers to steal credentials and hijack virtual machines

Insikt Group has reported , a previously undocumented command-and-control and post-exploitation framework for compromised hosts. It supports credential theft, virtual machine control and network intrusion. The framework requires existing root access, and the activity is linked to TAG-218, whose activity overlaps with .

Understand

Research, policy and context for the decisions ahead.

Security Affairs· · Critical severity

Google warns attackers are using autonomous AI agents to automate intrusions and scale exploitation

The Google Threat Intelligence Group reports that attackers are deploying autonomous multi-agent frameworks such as CAIRN and PentAGI to run multi-stage intrusions faster, and some are automating patch reverse engineering to weaponise n-day flaws. The techniques are familiar, relying on stolen credentials and unpatched systems. has handled LLMjacking incidents where fraudulent charges reached $1 million in the worst case before cleanup.

CSO Online·

Amazon reports criminals using ARTEX AI agents to steal AI credentials and bypass crypto exchange ID checks

Threat Intelligence says criminals ran campaigns using ARTEX, an open-source agentic AI penetration testing framework, with moderate confidence that several China-linked groups were involved. One campaign hunted for exposed API keys in AI gateways and development platforms such as , and RAGFlow. A second used real-time deepfakes and forged ID photos to try to defeat KYC checks at two crypto exchanges. Amazon has published the IP addresses seen between 14 October and 2 November so organisations can check their systems.

Cyber Security News·

Google reports attackers now using AI agents to automate real-world intrusions

and Mandiant report that threat actors are using open-source multi-agent hacking frameworks such as ARTEX, CAIRN and PentAGI for reconnaissance, privilege escalation and lateral movement. GTIG also found an exposed framework that found, wrote and ran exploits on its own, and reported an AI-assisted campaign against PaperCut flaws that reached domain admin within minutes. Separately, attackers are stealing AI keys and cloud access for resale, and fake charges have reached very high sums.

How priority and exploitation labels work · Report an error