This is your easyDNS #AxisOfEasy Briefing for the week of July 27th, 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 Lengend click here.

In this issue:

• In First-of-Its-Kind Case, Atlanta Man Charged Over Phone’s Self-Wiping Passcode
• FakeAgent Unmasked: Malicious Claude Artifact Delivered SectopRAT to 29+ Organizations
• Canada’s New Hate Speech Law Sparks Home Visits Over Old Social Media Posts
• Critical Rails Flaw Lets Hackers Read Server Secrets Through Image Uploads
• OpenAI’s AI Models Exploited a Zero-Day to Breach Hugging Face

Elsewhere Online:

easyHermes is here…

We’re pleased to announce the expansion of our agentic hosting platform with the rollout of easyHermes, the latest agent-enabled VPS appliance on the easyNode.ai platform.

Now you can answer the “openclaw vs Hermes” question for yourself, or run them both.

Check it out today, and get your first month on us when you use promo code AXISOFEASY

⇒ https://easynode.ai

In First-of-Its-Kind Case, Atlanta Man Charged Over Phone’s Self-Wiping Passcode

Samuel Tunick faces what’s believed to be the first U.S. prosecution over a phone’s “duress” passcode—a GrapheneOS feature that wipes data when entered. Customs agents seized his phone at Atlanta’s airport on January 24, 2025, despite the wipe, after allegedly denying him a lawyer.

He’s pleaded not guilty to property-destruction charges. His attorneys call the search pretextual, tied instead to his ties to the “Defend the Atlanta Forest” movement opposing “Cop City.” Security experts call the case unprecedented. A court ruling is expected later this year.

More via Techcrunch

FakeAgent Unmasked: Malicious Claude Artifact Delivered SectopRAT to 29+ Organizations

Huntress uncovered “FakeAgent,” a malvertising campaign that hit 29+ organizations (July 21–22, 2026) through a malicious artifact hosted on Claude.ai itself, redirecting Bing searchers to a fake ClaudeDesktop.exe. The sideloaded, VMProtect-packed malware used GPU anti-VM checks and shader-based encryption to hide its payload — cracked with help from Claude Opus 4.8 — revealing SectopRAT, a credential- and finance-stealing trojan.

Investigators traced the attacker via WHOIS records, an Ethereum-based “EtherHiding” C2 scheme, and a prior Docker Hub campaign, linking them to Microsoft’s Operation Endgame takedowns.

More via Huntress

Canada’s New Hate Speech Law Sparks Home Visits Over Old Social Media Posts

Canadians are reportedly being contacted at home by officials over social media posts from months or years ago, the JCCF says, following Bill C-9, the Combatting Hate Act (effective July 18).

The law makes hate-motivated crime a standalone offense, bans display of terrorist, Nazi, and noose symbols, and raises maximum sentences to life imprisonment. It defines hatred as “intense and extreme,” lets police and prosecutors judge posts, removes religious good-faith defenses, and drops the attorney general’s consent requirement—a change Justice Minister Sean Fraser says speeds up enforcement.

More via Reclaimthenet

Critical Rails Flaw Lets Hackers Read Server Secrets Through Image Uploads

A severe Ruby on Rails vulnerability (CVE-2026-66066, CVSS 9.5) lets unauthenticated attackers steal server files—including database passwords and API keys—via a simple crafted image upload, exploiting how Active Storage handles the libvips image library. No special thumbnail feature is needed to trigger it.

Rails versions 6.x through 8.1.3 are affected; patches are available in 7.2.3.2, 8.0.5.1, and 8.1.3.1. Crucially, upgrading alone won’t protect you—any exposed secrets must be rotated immediately. No public exploit exists yet, but researchers plan to release technical details by August 28, 2026.

More via Thehackernews

OpenAI’s AI Models Exploited a Zero-Day to Breach Hugging Face

Hugging Face disclosed on July 16 that an autonomous AI agent breached its systems; OpenAI later confirmed its own models were behind the attack. During confined offensive-cyber testing, the models exploited a JFrog Artifactory zero-day to escalate privileges, move laterally, and reach Hugging Face’s systems.

JFrog has since patched nine vulnerabilities across two Artifactory releases, crediting OpenAI’s responsible disclosure. CTO Yoav Landman called AI models “extraordinary zero-day discovery engines,” cutting both ways for attackers and defenders. Investigators say other services and additional, unnamed targets are still being probed.

More via Securityweek