Triggered silently when a user opens a crafted email message, the Outlook spoofing flaw leaked corporate NTLMv2 password hashes — and was uncovered with help from Anthropic’s Claude.
Researchers publish proof-of-concept exploit code that turns Microsoft’s own `.cnpf` log-authentication mitigation into a local privilege escalation vector on fully patched Windows systems.
Exploiting a world-writable local socket in the DynamicCLI daemon, ransomware and wiper crews pivot from network gateways into VMware ESXi hypervisors.
Unauthorized access into Japan’s largest furniture and home-goods chain compromises tens of millions of loyalty accounts during a week of intense MSP infrastructure alerts.
Built originally for red-team security audits, agentic penetration tools are being wired into uncensored LLMs to automate reconnaissance, exploit chaining, and data exfiltration.
The U.S. Department of Justice and international law enforcement seize more than 100 servers powering the Gootloader initial-access syndicate while major retail and aviation brands battle fresh ransomware extortion claims.
Attackers are chaining a Check Point Gaia OS zero-day (CVE-2026-123039) to gain root access and deploy destructive VMware ESXi wipers, as CSO disclosures confirm Anthropic’s Claude helped discover Outlook zero-day CVE-2026-124194.
Microsoft patched over 1,000 vulnerabilities including actively exploited Outlook zero-day CVE-2026-124194 alongside Google’s emergency Chrome XML patch (CVE-2026-125350) and the Fireworks AI / Notion AI "LLMJacking" breach.
Emergency patching continues after autonomous AI scanners found critical vulnerabilities in PHP—used by 70% of websites—while Microsoft Research demonstrates the Lazarus agent backdoor.
Classified as Critical under OpenAI’s Preparedness Framework, the autonomous security model also flagged a CVSS 9.2 vulnerability in Brotli as defenders race attackers.
The NCSC's annual report says AI-powered malware is highly likely to be used against national infrastructure within two years, and announces new defences.
What's next in AI.
The daily briefing on AI news, tools and breakthroughs — from every corner of the world.