Armadin researcher @mhskai2017 has uncovered two authenticated RCE vulnerability affecting Dell OpenManage Integration and Lenovo XClarity Integration with Windows Admin Center (CVE-2026-101207 and CVE-2026-65949, both High severity CVSS 8.8). A domain user credential is enough to gain SYSTEM on the WAC host, hijack admin sessions, and move laterally to every asset it manages. Patch immediately. Advisory and mitigation details: and

A week from today I’m presenting at @WWHackinFest! I’ll be sharing how offensive security operators can stay irreplaceable in the age of AI. We’ll cover our Application-First methodology and why humans still catch critical vulns that frontier LLMs miss.

Armadin raises $255.5M in Series B funding co-led by @a16z and @accel to bring proof, not assumptions, to the organizations that need it most. Read more:

Our latest kill chain is a spicy one! Two 0days and a path from the open internet straight to the corporate network. Armadin Director of Capabilities Nick McClendon joins VP of Product Management Greg Heon to walk you through the entire chain: decompiling a cloud backup server's Java application, an auth bypass via a secret value the app returned to its own front end, and a file upload to the web root that landed an undetected web shell running as SYSTEM on a host connected to Active Directory. It's a perfect example of why AI attacks routinely surface issues annual pen testing would never catch. It's also a reminder that tech debt can create real business risk in the AI era. YouTube: Spotify: Apple Podcast: Amazon Music: iHeart Radio:

Armadin researcher @0xc0ffee_ has uncovered a critical unauthenticated RCE vulnerability affecting all versions of SailPoint IdentityIQ (CVE-2026-12342). Given IdentityIQ handles sensitive credentials and access, this flaw presents a high risk of secret exposure and lateral movement into internal networks. Patch immediately. Advisory and mitigation details:

Kill Chains and Coffee is now available as a podcast. Listen on Apple, Spotify, or wherever you get your shows. Each episode walks a full attack path end to end: initial access, privesc, lateral movement, actions on objectives, plus where defenders can break the chain. 🎧 [links] - Spotify: - Apple Podcast: - Amazon Music: - iHeart Radio:

Human-scale security was never designed for machine-speed adversaries that can discover and exploit weaknesses faster than you can respond. How would your environment hold up against these AI-driven attacks? A Hyperattack™ assessment is how you find out. Tens of thousands of autonomous agents trained on elite human tradecraft. Finding every path across every domain, under strict safety controls. AI attacks are relentless. So are we. Know your risk before attackers do:

175 targets. 6,000 AI agents. 220,000 actions. One kill chain. In Ep 5 of "Kill Chains and Coffee," Craig Wright and Greg Heon break down a Hyperattack we ran for an organization with 22,000+ internet-facing services. Our agents targeted less than 1% of them. On one host, an agent read a route name and guessed it was a registration endpoint. No register button in the UI, but the API accepted the request. It created an account and got a token. From there: three SQL injections, linked servers into five more databases, and 68.5M rows of sensitive data, including E911 caller location and SIM-swap precursor data. The agent that registered the account passed it back, and the rest of the swarm took it from there. Different agents, different specialties, one chain. Watch:

In Episode 5 of "Kill Chains and Coffee," Craig Wright joins host Greg Heon to walk through a critical kill chain from a recent AI Hyperattack Armadin ran against a large enterprise. The scope was massive: 22,749 internet-facing services, with 5,724 agents taking 220,479 actions. Our AI attacker found a hidden registration endpoint, created its own account, and chained its way to SQL injection, exposing 68.5 million rows of workforce PII, 911 caller geolocation data, and SIM-swap precursors. Discover why an 80-hour pentest was never going to reach the 175th target, and why machine speed and scale are critical for offensive security in the AI era. Watch Episode 5: