Autonomous IT

Breaking from the normal Patch Tuesday cadence for an emergency drop. On May 7, security researcher Hyunwoo Kim published a working proof-of-concept for DirtyFrag - a Linux kernel local privilege escalation chain that gets unprivileged users to root on every major distribution. The embargo was broken by a third party before distribution backports were ready, so the exploit is public and the patch is not.

CTO Jason Kikta and Landon Miles walk through what makes DirtyFrag different from the Copy Fail mitigation many teams already deployed (spoiler: the CopyFail mitigation does NOT cover this), why AWS is calling it a class rather than a single CVE, and the five kernel modules you need to block right now: esp4, esp6, ipcomp4, ipcomp6, and rxrpc.

In this episode:

• Why the embargo break matters and what changed on May 7
• How DirtyFrag chains CVE-2026-43284 and CVE-2026-43500 to defeat both Ubuntu's namespace policy and the absence of rxrpc.ko on other distros
• Why this is the third generation of a bug class (DirtyPipe → Copy Fail → DirtyFrag) and what that means for what comes next
• The Automox Worklet that mitigates both arms across your Linux fleet, and what it deliberately does not do
• Tested affected platforms: Ubuntu 24.04, RHEL 10.1, AlmaLinux 10, CentOS Stream 10, openSUSE Tumbleweed, Fedora 44

Back to the regular Patch Tuesday schedule next week.

Links:

• Full blog post and mitigation guidance 
• Automox Worklet (in-console for customers): 
• Worklet source on GitHub
• Hyunwoo Kim's PoC and write-up
• AWS Security Bulletin 2026-027
• CVE-2026-31431 (Copy Fail, parent issue)

In this episode of Autonomous IT, Live!, we break down the widening gap between exploitation speed and remediation reality. Disclosed vulnerabilities keep climbing, exploitation windows keep shrinking, and IT and security teams are expected to absorb more risk without more resources. The traditional playbook — manual patching, fragmented workflows, scheduled cycles — was built for a slower world that no longer exists.

What you'll learn:

• Why threat actors consistently outpace defender response times
• Where manual patching and fragmented processes break down, even for mature teams
• How rising vulnerability volume and shrinking exploitation timelines are reshaping risk
• Why working harder isn't the answer — and what actually needs to change

Who should listen: IT and security leaders responsible for vulnerability management, infrastructure teams running distributed or SaaS-heavy environments, and anyone focused on shrinking exposure windows and accelerating response.

The gap between attacker speed and defender capability isn't closing on its own. This conversation is about what it takes to close it.

This live show originally aired April 22, 2026. 

Claude Mythos dominated the AI security conversation for two weeks straight, from the Cloud Security Alliance's strategy briefing to sharp public skepticism to yesterday's Bloomberg report that unauthorized users on Discord have been accessing Mythos since its limited launch. Host Jason Kikta cuts through the noise to separate the contested vendor claims from the established trend.

In this episode:

• Why the Mythos debate misses the point, and the independently verified AI security milestones that predate it (XBOW topping HackerOne, DARPA's AI Cyber Challenge, Google Big Sleep, Claude Opus 4.6's 500+ high-severity findings)
• A careful look at the numbers behind Anthropic's system card, including the Firefox exploit rate dropping from 72.4% to 4.4% once pre-discovered bugs are removed
• The CSA's top CISO recommendations that hold regardless of which Mythos claims you believe: patching, segmentation, egress filtering, MFA, defense in depth
• Three concrete actions to take this week, including the governance conversation most security leaders are overdue to have with the business

Good security starts with good IT. The trend is stable. The claims are contested. Anchor your planning accordingly.

Links and sources:

• CSA briefing
• Project Glasswing
• Mythos technical writeup
• Ottenheimer system card teardown
• Tom's Hardware on the 198 manual reviews: 
• Bloomberg on the Discord leak

This month's Patch Tuesday drops a SQL Server elevation of privilege that hands attackers sysadmin access and an actively exploited SharePoint XSS flaw that requires no authentication. 

SQL injection in the database engine. Cross-site scripting. In 2026...? 

Ryan and Mat break down how these attacks work, what to watch for, and why these "classic" vulnerability classes refuse to stay dead.                

Also covered: 80 Edge and Chromium fixes released this month, and a recurring reminder about Secure Boot certificates you can't afford to ignore this year.

In this episode of IT Insiders, Maddie Regis speaks with Emily Pace, a Senior Product Manager at Automox. Emily shares her career journey, her role in product management, and the collaborative environment at Automox that fosters innovation. They discuss current projects, the importance of customer feedback, and Emily's advice for IT professionals. The conversation concludes with a fun game about real and fake products, showcasing Emily's quick thinking and humor.

This episode originally aired November 19, 2024

What if your IT team could troubleshoot and remediate endpoint issues without ever leaving their service desk? In this episode, Steph Rizzuto and Katherine Chipdey break down the new Automox + Zendesk integration. 

They cover what it does, why it matters, and how it's designed to cut meantime to remediation for IT teams. The integration surfaces real-time endpoint data directly inside Zendesk tickets. It also gives admins one-click actions like device restarts, patch deployment, and policy execution. Beyond reactive fixes, it flags automation gaps so fewer tickets get created in the first place. Plus, hear how the Splashtop remote desktop integration ties it all together when hands-on troubleshooting is needed. Whether you're running a lean IT team or managing thousands of endpoints, this one's worth a listen.

In this episode of Automox Insiders, host Maddie Regis chats with Adam Whitman, Manager of Solutions Engineering at Automox, about all things IT spring cleaning. From patch management and software audits to business continuity planning and endpoint hygiene, Adam shares practical, real-world tips for tidying up your tech stack and staying ahead of IT clutter. 

Along the way, he reflects on his career journey from marketing to IT leadership and reveals some personal spring cleaning confessions. Tune in for expert advice and a fresh perspective to help you refresh your IT environment this season.

This podcast originally aired April 24, 2025

A routine support ticket reveals a deeper truth about patch management. Your patching tool says "done," but your vulnerability scanner disagrees – so who's right? In this episode, Automox SVP of Customer Experience Charles Coaxum walks through a real customer case. Dozens of machines showed as patched in the dashboard but were flagged as vulnerable by the scanner. The culprit wasn't a bug or misconfiguration. It was the often-overlooked gap between patch execution and patch completion. Learn why pending reboots create silent compliance gaps. Discover how to identify endpoints stuck in limbo. And see how automation turns hours of manual investigation into a minutes-long fix.

March 2026's Patch Tuesday brings no active exploitations, but don't let that fool you. This month, Ryan Braunstein and Henry Smith break down why medium-severity vulnerabilities deserve your full attention.

First up: a Push Message Routing Service memory leak (CVE-2026-24282, CVSS 5.5) that lets attackers scrape session tokens and private keys from heap memory. Then, a pair of GDI bugs (CVE-2026-25181 and CVE-2026-25190) that chain together to defeat ASLR and deliver remote code execution with near-perfect reliability. Henry covers a Windows Accessibility Infrastructure flaw (CVE-2026-24291) hiding in a service most teams never think to harden, plus an SMB authentication bypass (CVE-2026-24294) that echoes EternalBlue and WannaCry.

What you'll learn:

- How attackers chain medium-severity bugs into full compromise paths

- Why the Push Message Routing Service is a target-rich environment for credential theft

- How a two-stage GDI exploit defeats ASLR with near-100% reliability

- Why accessibility services are blind spots on your hardening checklists

- What SMB's history with EternalBlue and WannaCry means for this month's auth bypass

Patch your systems. Audit your service accounts. Don't skip the mediums.

In this episode, Jeremy Maldonado shares his experiences and insights on server management, highlighting the importance of learning from mistakes, the power of automation, and finding balance between Linux and Windows environments. He discusses the challenges and rewards of managing servers, the pivotal role of Ansible in streamlining operations, and the confidence required to maintain a reliable infrastructure. Jeremy encourages listeners to view setbacks as opportunities for growth while reminding us to be kind to ourselves throughout our professional journeys.

This episode originally aired March 6, 2025