Microsoft’s July 2026 Patch Tuesday fixed a record 570 flaws and three zero-days, two already exploited in AD FS and SharePoint identity infrastructure.

Qilin, formerly Agenda, was the most prolific ransomware-as-a-service of H1 2026 with 641 claimed victims, including the Asahi brewery attack.

In 48 hours the US, UK and EU indicted a bulletproof host and sanctioned VPN and cryptor sellers behind LockBit, Play and BlackSuit ransomware.

The Gentlemen ransomware weaponised a zero-day in Kontron’s ktapi.sys driver to gain kernel access and kill EDR from Microsoft, ESET, Palo Alto, and SentinelOne, researchers found.

Arctic Wolf says Anubis ransomware affiliates are exploiting Citrix Bleed 2 (CVE-2025-5777) and abusing legitimate RMM tools to breach networks, then deploying an irreversible data wiper.

A US government entity paid Kairos about $1 million to keep stolen files offline, a Ransom-ISAC case study shows. Kairos never encrypted a machine, it just threatened to publish.

Sysdig documented JADEPUFFER, the first ransomware attack whose technical execution was run end to end by an AI agent. A human handler still picked the target, built the infrastructure, and supplied credentials.

Inside XSS.is, the Russian cybercrime forum seized in 2025. A data-led profile from 123,241 leaked messages: what it traded, who ran it, its place in the ransomware kill chain, and a searchable country IoC table.

Agentic AI moves the threat from what a model says to what it does. We map how MCP turns goal hijacking, tool misuse, and privilege abuse into a working attack chain, and the controls that contain it.

A technical map of the Model Context Protocol attack surface in 2026: tool poisoning, line jumping, rug pulls, tool-chaining exfiltration, token sprawl, and the RCE flaws that turned MCP servers into entry points.