--- title: "#AxisOfEasy 474: Warhound Raises $200M To Engineer War Dogs And Arm Them With Anduril Gear" type: "post" post_id: "34510" slug: "axisofeasy-474-warhound-raises-200m-to-engineer-war-dogs-and-arm-them-with-anduril-gear" canonical: "https://axisofeasy.com/aoe/axisofeasy-474-warhound-raises-200m-to-engineer-war-dogs-and-arm-them-with-anduril-gear/" markdown_url: "https://axisofeasy.com/aoe/axisofeasy-474-warhound-raises-200m-to-engineer-war-dogs-and-arm-them-with-anduril-gear.md" json_url: "https://axisofeasy.com/aoe/axisofeasy-474-warhound-raises-200m-to-engineer-war-dogs-and-arm-them-with-anduril-gear.json" txt_url: "https://axisofeasy.com/aoe/axisofeasy-474-warhound-raises-200m-to-engineer-war-dogs-and-arm-them-with-anduril-gear.txt" published: "2026-10-30T22:00:48+00:00" modified: "2026-10-30T15:12:03+00:00" author: "Mark E. Jeftovic" categories: - "#AxisOfEasy" tags: excerpt: "Warhound Raises $200M To Engineer War Dogs And Arm Them With Anduril Gear An MCP That Reads Your Reddits Quietly Runs Hidden Python Stuffed Into An Image Japan's Onamae Desktop Cloud Hacked, Malware On Up To 27,000 Servers... this and more in AofE #474" site_name: "" publisher: "" language: "en-US" license: "" generator: "easyPress AI Discoverability" generator_version: "2.0.14" --- *![](https://axisofeasy.com/wp-content/uploads/2020/02/weekly-briefing-axis-of-easy.jpg)* --- *Weekly Axis Of Easy #474* ========================== --- ***Last Week’s Quote was:** “No man is free who is not master of himself.” Solved — Epictetus! Congratulations Veronika Ho for taking the prize this round.* ***This Week’s Quote:** “We choose mania over boredom every time.” By ???* ***THE RULES: No searching up the answer,** must be posted at the bottom of this blog post, in the comments section.* ***The Prize:** First person to post the correct answer gets their next domain or hosting renewal on us.* --- *This is your easyDNS #AxisOfEasy Briefing for the week of October 26th, 2026. Our Technology Correspondent Joann L Barnes and easyCEO Mark E. Jeftovic send out a short briefing on the state of the ‘net and how it affects your business, security and privacy.* *To Listen/watch this podcast edition with commentary and insight from Joey and Len the Legend [click here.](https://axisofeasy.com/podcast/axisofeasy-474-warhound-raises-200m-to-engineer-war-dogs-and-arm-them-with-anduril-gear/)* **Is Your Website Vulnerable to PHP2shell? (and what else)** We’ve added a PHP2shell scanner to [DomainHelp here](https://mc.easydns.com/r/aabe1142b472093bfdc3e7673?ct=YTo1OntzOjY6InNvdXJjZSI7YToyOntpOjA7czo1OiJlbWFpbCI7aToxO2k6MjAwNzt9czo1OiJlbWFpbCI7aToyMDA3O3M6NDoic3RhdCI7czoyMjoiNmFkYjliZjQ3MTA4ZDY1NzA1NDQzMCI7czo0OiJsZWFkIjtzOjU6IjE5NTUyIjtzOjc6ImNoYW5uZWwiO2E6MTp7czo1OiJlbWFpbCI7aToyMDA3O319&) which can tell you whether your website is vulnerable. It is based on Assetnote’s [Git repo](https://mc.easydns.com/r/05ed3d5af31b46134d90d2e6e?ct=YTo1OntzOjY6InNvdXJjZSI7YToyOntpOjA7czo1OiJlbWFpbCI7aToxO2k6MjAwNzt9czo1OiJlbWFpbCI7aToyMDA3O3M6NDoic3RhdCI7czoyMjoiNmFkYjliZjQ3MTA4ZDY1NzA1NDQzMCI7czo0OiJsZWFkIjtzOjU6IjE5NTUyIjtzOjc6ImNoYW5uZWwiO2E6MTp7czo1OiJlbWFpbCI7aToyMDA3O319&) (be sure to give him a star when you clone it). **PHP2Shell** is just the beginning – we’ve scored a deal for AxisOfEasy readers on a comprehensive “after *after* PHP2shell” that you should be worried about (including the new Node.js remote shell! Huzzah). Use coupon code “AXISOFEASY” [order your vulnerability report here:](https://mc.easydns.com/r/540b6ae6dd0aeeba9dce4bbe5?ct=YTo1OntzOjY6InNvdXJjZSI7YToyOntpOjA7czo1OiJlbWFpbCI7aToxO2k6MjAwNzt9czo1OiJlbWFpbCI7aToyMDA3O3M6NDoic3RhdCI7czoyMjoiNmFkYjliZjQ3MTA4ZDY1NzA1NDQzMCI7czo0OiJsZWFkIjtzOjU6IjE5NTUyIjtzOjc6ImNoYW5uZWwiO2E6MTp7czo1OiJlbWFpbCI7aToyMDA3O319&) [![](https://axisofeasy.com/wp-content/uploads/2026/10/axisofeasy473add.png)](https://zybee.ai/exposure-report/?coupon=AXISOFEASY) ***In this issue:*** - **Warhound Raises $200M To Engineer War Dogs And Arm Them With Anduril Gear** - **An MCP That Reads Your Reddits Quietly Runs Hidden Python Stuffed Into An Image** - **Japan’s Onamae Desktop Cloud Hacked, Malware On Up To 27,000 Servers** - **Tailscale Warns of 11 Unpatched SSH Flaws Allowing Root Privilege Escalation** - **Anthropic’s Unreleased AI Model Chains Flaws Into Working Chrome, Safari, Windows and Linux Exploits** ***Elsewhere Online:*** - **Sony-Owned Internet Provider So-Net Breach Exposes 2.6M Customer Records** - **University Of Copenhagen Data Breach Exposes 525,000 Records** - **Energy Giant ENGIE Discloses Customer Data Breach** - **Lazarus Group Abuses South-Korean Security Software To Deploy Post-Quantum-Encrypted Malware** - **Australia Cuts Off 180,000 Infected Computers Via Amadey Malware Backend Take-Down** --- [**Warhound Raises $200M To Engineer War Dogs And Arm Them With Anduril Gear**](https://techcrunch.com/2026/10/22/ben-lamm-raises-200m-for-warhound-to-engineer-war-dogs-and-arm-them-with-anduril-gear/) Ben Lamm and Palmer Luckey’s company Warhound has emerged from stealth with a $200M raise to genetically engineer working dogs for the military and law-enforcement market — the dogs will take on bomb detection, searches and other jobs that are dangerous for humans or awkward for machines. The premise is a supply problem: more than 80% of the U.S. military’s working dogs come from Europe, the domestic pipeline keeps thinning, and many of the top breeds (Belgian Malinois, German Shepherd) fail training over inherited predispositions. Lamm’s plan is to clone for health, longevity, speed and a sharper nose; Colossal — best known for de-extinction work — licenses the gene technology to the venture, and its sister Viagen will run the cloning. The first batch is being prepped now, with births expected early next year and service within two years. Anduril supplies the gear “on top of the dog, not in the dog” — sensor vests, goggles and other kit that give the handler a real-time situational-awareness feed — while the sales target runs to first responders, the Department of Defense and police departments. The practical close: “genetically engineered war dog” is now a live vendor claim in a defense-procurement context, so if you’re on the buying side that’s a new line item to vet — cloning success rates, welfare provisions, and service warranties — that doesn’t show up in the usual defense-supply questionnaires. More via [TechCrunch](https://techcrunch.com/2026/10/22/ben-lamm-raises-200m-for-warhound-to-engineer-war-dogs-and-arm-them-with-anduril-gear/) · [Wired](https://www.wired.com/story/cool-cool-cool-palmer-luckey-made-a-genetically-modified-military-dog-startup/) [**An MCP That Reads Your Reddits Quietly Runs Hidden Python Stuffed Into An Image**](https://www.knostic.ai/blog/mcp-that-reads-your-reddits-and-runs-hidden-python-in-an-image) A PyPI package that exposes a single tool — reddit\_extract, which takes a Reddit URL and returns the post plus up to 50 comments — was quietly running hidden Python pulled out of a PNG image every time it was called. Knostic’s write-up dissects the mechanism: on each call the package downloads an image from a fixed URL, reads the lowest bit of each RGB channel across the pixels, and zlib-decodes the result into runnable Python. The image host is effectively the code host — whatever the URL serves at the moment of the call is what runs on the caller’s machine — and there are no install-time hooks at all; the execution path is entirely per-call, gated only by a successful Reddit fetch, a successful image download, and a clean decode. The image in Knostic’s capture was a proof-of-execution demo — no exfiltration or persistence — and the researchers traced the code back to the documented educational proof-of-concept in a public ethical-hacking open-source repository. The trap was that the PyPI package (uploaded February 7, 2026 as version 0.1.0) shipped without the parent repo’s README explaining what it was doing, so an installer sees “a reddit tool” — not “a stealth-execution PoC.” Practical close: if your agents run third-party MCPs or PyPI packages, treat the install step as the least interesting surface — audit the package’s network calls and anything it downloads at runtime, because a per-call fetch to a host-controlled URL is exactly the seam that turns an innocent tool into a remote-code path. *More via [Knostic](https://www.knostic.ai/blog/mcp-that-reads-your-reddits-and-runs-hidden-python-in-an-image)* [**Japan’s Onamae Desktop Cloud Hacked, Malware On Up To 27,000 Servers**](https://www.bleepingcomputer.com/news/security/japans-onamae-desktop-cloud-hacked-up-to-27-000-servers-affected/) The infrastructure behind Onamae Desktop Cloud — a Japanese VPS service operated by GMO Internet and marketed heavily toward FX automated trading and continuous MetaTrader 4 runs — was breached, and the company says “malicious programs” were installed on up to 26,739 customer servers. Onamae confirmed on October 20 that a third party had entered the service, and began progressively shutting customer servers down starting at 04:00 local time on October 21, with the headline affected count rounding out to “up to 27,000.” The company says data stored on the impacted servers may have been exposed, but card and payment details are out of scope — they’re held on a separate stack. Practical close: if you ran credentials, API keys, or bot state on this service in the affected window, treat them as burned and rotate; more broadly, running an FX/MT4 bot farm on shared managed VPS infrastructure means an infra-level compromise inside the provider is precisely the blast radius that model was designed to reach — so pin whatever you can to isolated single-tenant hardware, and treat any “VPS” that’s really shared host as shared-tenant by definition. *More via [BleepingComputer](https://www.bleepingcomputer.com/news/security/japans-onamae-desktop-cloud-hacked-up-to-27-000-servers-affected/)* [**Tailscale Warns of 11 Unpatched SSH Flaws Allowing Root Privilege Escalation**](https://www.bleepingcomputer.com/news/security/tailscale-warns-of-serious-ssh-flaws-allowing-root-privilege-escalation/) Cyberdione Labs disclosed Tailworm, a collection of 11 unpatched flaws found across Tailscale 1.104.2 and the self-hosted headscale v0.29.5 control server. The most severe is an SSH root-shell path: an attacker with any valid account on a target in the tailnet abuses a file-creation primitive in Tailscale’s SSH socket forwarding to plant a file that tailscaled later treats as a startup script, gaining code execution inside the daemon — which runs as root. The remaining ten flaws cover credential theft, DoS, and a path that writes attacker-crafted JSON into headscale’s database, allowing an attacker to register a new machine in a victim’s tailnet and then reach it by IP. The team says all were reproduced in a lab on unmodified, released software, and Tailscale confirms it is working on patches. Practical close: if you run tailscale serve/ssh or headscale in production, treat any account on the tailnet as a root-adjacent trust anchor — audit who and what can log in, and watch for the 1.104.3+ patch that lands this window. *More via [BleepingComputer](https://www.bleepingcomputer.com/news/security/tailscale-warns-of-serious-ssh-flaws-allowing-root-privilege-escalation/) · [SecurityLabz](https://www.securitylabz.com/cyberdione-discovers-11-tailscale-security-vulnerabilities-including-ssh-root-shell-chain/)* [**Anthropic’s Unreleased AI Model Chains Flaws Into Working Chrome, Safari, Windows and Linux Exploits**](https://thehackernews.com/2026/10/anthropics-unreleased-ai-model-chains.html) Anthropic’s Frontier Red Team used an unreleased internal model to find and chain multiple browser, OS, and mobile flaws into five working exploit chains for Chrome, Safari, Windows and Linux — reaching sandbox escape or cross-site data theft from a single page load with “little human involvement.” The strongest claim isn’t any single bug but that one model combined several separate flaws into a working exploit — work Anthropic says has until now been the job of a few specialist teams, one of which opened the Calculator app on a stock Mac running Chrome seconds after a page loaded. Anthropic reported 62 flaws to Google and Apple; both vendors credit Anthropic and its Claude model for specific fixes, and Chrome 156 (Oct 20) and iOS/iPadOS 26.5.2 (June 29) are already out with patches. What a reader can’t verify is how much of each chain the model built on its own — Anthropic doesn’t break down what a person did at each step or name the model beyond “unreleased, internal research.” Practical close: the threat here is not one CVE — it’s an agent that treats the whole attack surface as a single search problem and assembles the chain on its own; if you run a fleet of endpoints or a browser farm, “AI-generated exploit chain” is now a real risk category to add to your threat model. *More via [The Hacker News](https://thehackernews.com/2026/10/anthropics-unreleased-ai-model-chains.html)* **Also on AxisOfEasy this week:** *Nothing new in Curated Posts since our last check — still nothing since May. We’re not going to pad this section with filler just to fill space; when something worth your time crosses our desk, it’ll be here.* --- ***Elsewhere Online:*** **Sony-Owned Internet Provider So-Net Breach Exposes 2.6M Customer Records** *Read: https://www.helpnetsecurity.com/2026/10/28/sony-network-communications-data-breach-so-net/* **University Of Copenhagen Data Breach Exposes 525,000 Records** *Read: https://www.bleepingcomputer.com/news/security/university-of-copenhagen-data-breach-exposes-data-of-525-000-people/* **Energy Giant ENGIE Discloses Customer Data Breach** *Read: https://www.bleepingcomputer.com/news/security/energy-giant-engie-discloses-data-breach-affecting-customers/* **Lazarus Group Abuses South-Korean Security Software To Deploy Post-Quantum-Encrypted Malware** *Read: https://thehackernews.com/2026/10/lazarus-abuses-south-korean-security.html* **Australia Cuts Off 180,000 Infected Computers Via Amadey Malware Backend Take-Down** *Read: https://thehackernews.com/2026/10/australian-takedown-disrupts-amadey-botnet.html* --- **Previously on #AxisOfEasy** If you missed the previous issues, they can be read online here: - - - [October 23, 2026: #AxisOfEasy 473: PHP Ships Emergency Security Release For Two Critical RCE Flaws](https://axisofeasy.com/aoe/axisofeasy-473-php-ships-emergency-security-release-for-two-critical-rce-flaws/) - [October 16, 2026: #AxisOfEasy 472: A Flock-Like Camera In Canada, Pointed At The Police Station — Then The Police Visited](https://axisofeasy.com/472) - [October 9, 2026: #AxisOfEasy 471: Bitget Exchange Loses $387.5 Million After Attackers Exploit A Zero-Day In Security Software](https://axisofeasy.com/471) - [October 2, 2026: #AxisOfEasy 470: Citrix’s NetScaler Zero-Days Were Exploited For Weeks Before Anyone Patched](https://axisofeasy.com/470) - [September 25, 2026: #AxisOfEasy 469: Google Confirms Gemini Broke Into Three Real Companies During A Security Test](https://axisofeasy.com/469)