⸻

📄 Show Notes

🚨 CVE of the Week: Adobe Acrobat Zero-Day (CVE-2026-34621)

This week’s vulnerability is about as bad—and as common—as it gets. A zero-day flaw in Adobe Acrobat Reader is actively being exploited in the wild, requiring nothing more than opening a malicious PDF to trigger a full system compromise.

🔍 What Happened

•CVE ID: CVE-2026-34621

•Type: Zero-day (actively exploited before patch release)

•Severity: CVSS 8.6 (High, but misleading in practice)

•Attack Vector: Malicious PDF file

•Impact: Remote Code Execution (RCE), data theft

Adobe issued an emergency out-of-band patch, signaling the urgency and severity of the threat.

⸻

⚠️ Why This Is So Dangerous

This exploit is particularly concerning because:

•No user interaction required beyond opening a file

•Works through phishing and email attachments

•Targets one of the most widely used enterprise tools (PDF readers with ~60–75% market share)

Once triggered, the vulnerability exploits a memory corruption flaw (e.g., use-after-free or buffer overflow), allowing attackers to execute arbitrary code on the system.

⸻

🔗 The Real Threat: Exploit Chaining

On its own, this vulnerability is severe—but in modern environments, it’s even worse:

•Attackers use phishing to deliver the malicious PDF

•Gain access to a user endpoint

•Pivot into:

•Cloud infrastructure

•Container environments

•Internal systems

👉 This is how a “medium-high” CVSS score becomes a critical enterprise breach

⸻

🤖 AI and the Acceleration of Attacks

The pace of exploitation is changing:

•Exploits are now being weaponized within minutes of disclosure

•Attackers can deploy automated agents at scale

•AI-driven reconnaissance reduces time-to-exploit dramatically

This creates a world where patch latency = exposure window.

⸻

🛠️ Mitigation & Recommendations

Immediate Actions:

•✅ Patch Adobe Acrobat immediately (no delay)

•🚫 Do NOT wait for standard patch cycles

•📧 Treat all PDF attachments as potential attack vectors

Enterprise IT Best Practices:

•Enforce auto-updates and forced patching policies

•Consider network access restrictions for unpatched devices

•Implement:

•Zero Trust architectures

•Endpoint monitoring and anomaly detection

⸻

🧠 Strategic Takeaways

•User behavior is still the weakest link

•Patch cycles must shift from scheduled → real-time response

•Vendors must improve update mechanisms:

•Fewer forced reboots

•Better “do not interrupt” intelligence

We are entering a phase where patching speed is a primary security control, not a maintenance task.

⸻

💬 Listener Feedback

Thanks to listener IAPX for pointing out a technical clarification from last week:

•The Docker vulnerability discussed was rooted in Moby, not Docker directly

•Docker remains the primary exposure vector due to its widespread use

Great catch—and exactly the kind of feedback we appreciate.

⸻

📣 Wrap Up

Have thoughts on this vulnerability? Are we underestimating the impact of PDF-based attacks?

📧 Email: feedback@itsparccast.com

🐦 X: @itsparccast

💬 YouTube: Drop a comment—we read them all

⸻

🔗 Social Links

IT SPARC Cast

@ITSPARCCast on X

https://www.linkedin.com/company/sparc-sales/ on LinkedIn

John Barger

@JohnBarger on X

https://www.linkedin.com/in/johnbarger/ on LinkedIn

Lou Schmidt

@loudoggeek on X

https://www.linkedin.com/in/louis-schmidt-b102446/ on LinkedIn

Hosted on Acast. See acast.com/privacy for more information.

⸻

📄 Show Notes

🚨 CVE of the Week: Docker API Authorization Bypass (CVE-2026-34040)

This week’s CVE highlights a serious vulnerability in Docker Engine that undermines one of the core assumptions of container security: isolation.

🔍 What Happened

•CVE ID: CVE-2026-34040

•CVSS Score: 8.8 (High)

•Affected Systems: Docker Engine / Moby versions prior to 29.3.1

•Root Cause: Improper handling of authorization plugin checks in Docker’s API layer

The vulnerability allows specially crafted API requests to bypass authorization controls by dropping the request body before inspection—while still executing the request.

⸻

⚠️ Why This Matters

This flaw enables attackers to:

•Bypass container security policies

•Create privileged containers

•Access the host file system

•Extract sensitive credentials (SSH keys, cloud keys, etc.)

This effectively breaks container isolation, turning Docker from a security boundary into an attack vector.

⸻

🔗 The Bigger Risk: Chained Attacks

While Docker APIs are typically not exposed publicly, this vulnerability becomes significantly more dangerous in real-world environments:

•Attackers gain initial access via:

•Phishing or spear phishing

•Compromised endpoints

•Malware or trojans

•Then pivot internally to exploit Docker APIs

👉 In these scenarios, the practical severity approaches 9.8–10.0, not 8.8.

⸻

🤖 AI-Driven Threat Amplification

Modern attack frameworks—especially those leveraging AI—can:

•Automatically scan for exposed APIs

•Execute chained exploits without human intervention

•Scale attacks across thousands of targets simultaneously

This dramatically reduces the skill barrier for attackers.

⸻

🛠️ Mitigation & Recommendations

Immediate Actions:

•✅ Upgrade Docker to version 29.3.1 or later

•🔒 Restrict and lock down Docker API access

•🚫 Ensure APIs are not externally exposed

Strategic Recommendations:

•Enable auto-updates where operationally safe

•Conduct a full network audit (hosts, containers, firmware, network gear)

•Patch beyond servers:

•BIOS / firmware

•Network infrastructure (switches, routers)

•Break down silos between:

•Enterprise IT security

•Data center / cloud security

⸻

🔄 Key Takeaway

Containerization is not a silver bullet for security. Misconfigurations and API exposure can turn Docker into a high-impact attack surface—especially when combined with modern, automated attack chains.

⸻

💬 Listener Feedback

Thanks to listener PutlerLXO for correcting last week’s Axios stat:

•Actual weekly downloads: 100 million, not 45 million

We appreciate the feedback—keep it coming!

⸻

📣 Wrap Up

Have thoughts on this vulnerability? Think it’s overblown—or even worse than we described?

📧 Email: feedback@itsparccast.com

🐦 X: @itsparccast

💬 YouTube & LinkedIn: Drop a comment—we read them all

⸻

🔗 Social Links

IT SPARC Cast

@ITSPARCCast on X

https://www.linkedin.com/company/sparc-sales/ on LinkedIn

John Barger

@john_Video on X

https://www.linkedin.com/in/johnbarger/ on LinkedIn

Lou Schmidt

@loudoggeek on X

https://www.linkedin.com/in/louis-schmidt-b102446/ on LinkedIn

Hosted on Acast. See acast.com/privacy for more information.

Are tech CEOs using AI as cover for layoffs? Are emergency patches from major vendors signaling deeper systemic risk? And what’s really behind OpenAI’s decision to shut down Sora?

Plus, listener feedback sparks a deep dive into home router security and the best options for every level—from plug-and-play to prosumer setups.

If you’re in enterprise IT, security, or just trying to stay ahead of the curve, this is your weekly signal through the noise.

⸻

📌 Show Notes

00:00 – Intro

•Overview of the week’s biggest enterprise IT stories

•AI layoffs, patch failures, and shifting priorities in AI platforms

⸻

📰 News Bytes

00:49 – Tech CEOs Suddenly Love Blaming AI for Mass Job Cuts

•Increasing trend: layoffs attributed to “AI efficiency gains”

•Reality check: cost-cutting, restructuring, and execution failures

•Market dynamics:

•“AI-driven efficiency” messaging can stabilize or boost stock prices

•Traditional layoffs often trigger negative investor reactions

•Key takeaway:

•AI is becoming a narrative shield for leadership decisions

•Career insight:

•Job security = being a problem solver, not just a role filler

•Enterprise angle:

•Evaluate vendor stability when layoffs are framed as “AI transformation”

https://www.bbc.com/news/articles/cde5y2x51y8o

⸻

07:06 – Emergency Microsoft & Oracle Patches Point to Wider Cyber Issues

•Rise in out-of-band (emergency) patching

•Key incidents:

•Critical remote code execution vulnerability (CVSS 9.8)

•Broken update causing login failures

•Core issue:

•Patch reliability vs. urgency tradeoff is collapsing

•Enterprise implications:

•Traditional patch windows are becoming obsolete

•Delayed patching = increased exposure risk

•New reality:

•Mandatory, rapid patch deployment is now required

•Strategic shift:

•Move toward live patching architectures (already common in Linux/cloud)

•Root causes:

•Faster release cycles

•Increased reliance on automation

•Reduced staffing depth

https://www.computerweekly.com/news/366640648/Emergency-Microsoft-Oracle-patches-point-to-wider-cyber-issues

⸻

13:28 – Why OpenAI Really Shut Down Sora

•Contrary to speculation: not a collapse signal

•Actual drivers:

•Compute constraints

•Resource prioritization

•Revenue alignment

•Market dynamics:

•AI arms race: speed, capability, and scale

•Product reality:

•Video generation = extremely compute-intensive

•Limited sustained user demand vs. cost

•Strategic takeaway:

•Focus shifting toward:

•Coding tools

•Agentic platforms

•High-ROI capabilities

•Key insight:

•AI growth is currently compute-bound, not idea-bound

https://techcrunch.com/2026/03/29/why-openai-really-shut-down-sora/

⸻

📬 16:54 – Mail Bag & Home Router Recommendations

Listener Feedback Topics:

•Router security concerns

•Safer alternatives to high-risk vendors

Recommended Router Tiers:

🟢 Entry-Level (Simple / Plug-and-Play)

•Netgear

•Strong open-source firmware support (OpenWRT, Tomato)

•U.S.-based company with supply chain flexibility

•High accountability and responsiveness

🟡 Mid-Tier (Mesh / Larger Homes)

•Eero (Amazon-owned)

•Strong performance and ease of use

•Consistent updates and long-term viability

🔵 Prosumer / Advanced

•Ubiquiti (UniFi)

•Best-in-class price/performance

•Full ecosystem: networking + security + cameras

•No recurring cloud fees

•Strong automation and patch responsiveness

⸻

🔚 26:54 – Wrap Up

•Call for listener feedback

•Engage via email, X, YouTube, or LinkedIn

•Reminder to like, subscribe, and enable notifications

⸻

🌐 Social Links

IT SPARC Cast

@ITSPARCCast on X

https://www.linkedin.com/company/sparc-sales/ on LinkedIn

John Barger

@john_Video on X

https://www.linkedin.com/in/johnbarger/ on LinkedIn

Lou Schmidt

@loudoggeek on X

https://www.linkedin.com/in/louis-schmidt-b102446/ on LinkedIn

Hosted on Acast. See acast.com/privacy for more information.

Attackers compromised a maintainer account and injected malicious code into widely distributed versions, turning routine installs into a cross-platform Remote Access Trojan (RAT) deployment.

This isn’t just another vulnerability — it’s a breach of trust in the open-source ecosystem that powers modern web applications.

⸻

📝 Show Notes 

A major supply chain attack has compromised Axios, a core JavaScript library used in millions of applications across web, mobile, and backend systems.

In this episode of IT SPARC Cast – CVE of the Week, John Barger and Lou Schmidt explain how attackers injected malware into trusted Axios packages — impacting potentially tens of millions of environments worldwide.

⸻

🔎 What Happened

Axios is a widely used open-source library for making HTTP requests in:

•Node.js applications

•React, Angular, and Vue frontends

•Mobile apps (React Native)

•SaaS platforms and internal tools

With over 45 million weekly downloads, its footprint is enormous.

Attackers compromised an Axios maintainer’s NPM account and pushed malicious versions:

•Axios 1.14.1

•Axios 0.30.4

These versions introduced a hidden dependency:

•plain-crypto-js@4.2.1

This dependency executed a post-install script that deployed a cross-platform Remote Access Trojan (RAT) targeting:

•Windows

•macOS

•Linux

The malware then:

•Contacted a command-and-control (C2) server

•Downloaded OS-specific payloads

•Executed silently

•Deleted itself and restored clean package files to evade detection

⸻

⚠ Why This Is So Dangerous

This attack is particularly severe because:

•It does not require direct user action beyond installing dependencies

•It affects transitive dependencies (you may be using Axios without knowing it)

•It operates during build/install processes (CI/CD pipelines included)

•It leaves minimal forensic evidence

This is a classic supply chain compromise — not a CVE, but arguably more dangerous.

⸻

🏢 Enterprise IT Impact

If your organization:

•Uses Node.js or modern JavaScript frameworks

•Runs CI/CD pipelines

•Builds or deploys SaaS platforms

•Uses third-party APIs or SDKs

You are likely exposed.

Even if you don’t directly install Axios, it may exist deep in your dependency tree.

⸻

🧠 Key Takeaway

This was not a flaw in code.

This was a failure of trust in the supply chain.

If your security model assumes dependencies are safe by default — this attack proves otherwise.

⸻

🔗 Source Articles

https://thehackernews.com/2026/03/axios-supply-chain-attack-pushes-cross.html

https://www.elastic.co/security-labs/axios-supply-chain-compromise-detections

⸻

🔗 Connect With Us

IT SPARC Cast

@ITSPARCCast on X

https://www.linkedin.com/company/sparc-sales/ on LinkedIn

John Barger

@john_Video on X

https://www.linkedin.com/in/johnbarger/ on LinkedIn

Lou Schmidt

@loudoggeek on X

https://www.linkedin.com/in/louis-schmidt-b102446/ on LinkedIn

Hosted on Acast. See acast.com/privacy for more information.

Elon Musk announces TeraFab, a massive new effort to bring chip fabrication back in-house for greater control over AI hardware and supply chains. Mark Zuckerberg pushes deeper into agentic AI with plans for a personal “AI CEO” to manage workflows and decision-making. And Arm signals a major strategic shift with a new AI-focused chip designed for agent-based systems—putting it in direct competition with its own ecosystem.

From supply chain control and custom silicon to AI-driven leadership tools and next-generation chip architectures, this episode explores how the foundation of enterprise IT is rapidly evolving.  

⸻

⏱️ Show Notes

00:00 – Intro

📰 News Bytes

00:45 – Elon Musk Announces TeraFab for AI Chips and Memory

Elon Musk has announced plans for TeraFab, a massive chip fabrication initiative aimed at regaining full control over chip design and production.

The strategy includes:

• A prototype fabrication facility for rapid iteration

• A large-scale production fab for mass manufacturing

• Vertical integration to reduce dependency on external foundries

• Faster time-to-market for AI-driven hardware

As chip demand surges due to AI workloads, companies are reconsidering outsourced manufacturing models. TeraFab represents a return to end-to-end control of silicon development, which could significantly impact supply chains, pricing, and innovation speed.

https://x.com/i/broadcasts/1yKAPMzlvgWxb 

https://en.wikipedia.org/wiki/Terafab 

09:46 – Mark Zuckerberg Builds AI CEO to Help Run Meta

Mark Zuckerberg is developing a personal AI system capable of handling executive-level tasks—effectively functioning as a digital chief of staff or “AI CEO.”

The system is designed to:

• Retrieve and synthesize information across internal systems

• Automate decision-support workflows

• Reduce reliance on layers of management

• Act as a “second brain” for operational awareness

This reflects a broader shift toward agentic AI, where intelligent systems proactively execute tasks rather than simply responding to prompts. The discussion also raises key enterprise questions around security, portability, and ownership of personal AI agents.

https://www.the-independent.com/tech/mark-zuckerberg-ai-ceo-bot-b2943792.html

17:54 – Arm Unveils New AI Chip for Agentic Systems

Arm has announced a new AI-focused chip architecture aimed at powering agentic AI and future AGI-style workloads.

Key implications include:

• A shift from IP licensing to direct chip competition

• Increased competition with existing ecosystem partners

• Potential acceleration of specialized AI hardware development

• Growing relevance of alternative architectures like RISC-V

This move signals a major strategic pivot for Arm, potentially reshaping the competitive landscape for AI infrastructure and creating new dynamics between chip designers, manufacturers, and enterprise buyers.

https://www.reuters.com/business/media-telecom/arm-unveils-new-ai-chip-expects-it-add-billions-annual-revenue-2026-03-24/ 

🔁 Wrap Up

25:24 – Mail Bag

Listener feedback highlights continued interest in emerging compute models, including biological computing, and reinforces the importance of staying ahead of major infrastructure trends.

27:01 – Wrap Up

John and Lou close with thoughts on the convergence of AI, custom silicon, and agent-based workflows, emphasizing that enterprise IT leaders must prepare for a future where infrastructure, software, and decision-making are increasingly intertwined.

⸻

🔗 Connect With Us

IT SPARC Cast

@ITSPARCCast on X

https://www.linkedin.com/company/sparc-sales/ on LinkedIn

John Barger

@john_Video on X

https://www.linkedin.com/in/johnbarger/ on LinkedIn

Lou Schmidt

@loudoggeek on X

https://www.linkedin.com/in/louis-schmidt-b102446/ on LinkedIn

Hosted on Acast. See acast.com/privacy for more information.

From botnets built on consumer routers to concerns about firmware, silicon-level vulnerabilities, and manufacturing visibility, the conversation explores why your home or small office router may be one of the weakest links in modern cybersecurity.

The hosts explain what’s changing in the router market, which vendors are most at risk, and what both consumers and enterprise IT professionals should be doing now to secure the network edge.

⸻

📝 Show Notes

Consumer routers are no longer just simple networking devices — they are now prime targets in large-scale cyberattacks and botnet operations.

In this episode of IT SPARC Cast – CVE of the Week, John Barger and Lou Schmidt break down the growing risks tied to router supply chains, firmware security, and edge network vulnerabilities.

Rather than focusing on a single CVE, this discussion highlights a broader shift in how attackers are targeting home routers, small office devices, and prosumer networking gear as entry points into larger networks.

⸻

🔎 What’s Changing in Router Security

Recent attack trends show:

•Consumer and small-office routers are being used as launch points for larger cyberattacks

•Botnets are increasingly built on unpatched or poorly secured edge devices

•Attackers are leveraging routers to mask origin and evade detection

This makes routers one of the most critical — and often overlooked — components of modern security architecture.

⸻

⚠ The Supply Chain Problem

One of the biggest concerns discussed in this episode is supply chain visibility.

Key risks include:

•Limited insight into where hardware components are manufactured

•Potential for firmware-level or silicon-level vulnerabilities

•Difficulty auditing third-party manufacturing processes

•Inability to fully validate device integrity

Even when running trusted software (such as open-source firmware), underlying hardware risks may still exist.

⸻

🏢 Enterprise & Home Network Impact

This is not just a consumer issue.

Organizations must consider:

•Remote employees connecting via insecure home routers

•Small offices using low-cost networking equipment

•IoT devices relying on consumer-grade infrastructure

•Edge devices acting as entry points for lateral movement

If the edge is compromised, the rest of the network is exposed.

⸻

🛠 What IT Teams and Consumers Should Do

•Avoid default configurations and credentials

•Keep firmware updated consistently

•Segment home and corporate network traffic where possible

•Evaluate router vendors for security posture and supply chain transparency

•Monitor for unusual traffic patterns or device behavior

•Plan for longer-term shifts in router procurement and standards

This is a long-term evolution, not a short-term panic event.

⸻

📊 Market Impact & Vendor Landscape

The episode also discusses potential market shifts:

•Lower-cost vendors may face increased scrutiny

•Vendors with stronger supply chain transparency may benefit

•Manufacturing may shift to more trusted and auditable environments

•Future devices may require mandatory security features like auto-updating firmware

⸻

💬 Listener Feedback

Listener feedback from X highlights the growing importance of Zero Trust and identity validation, especially in response to recent discussions about insider threats.

The takeaway:

Security is no longer just about devices — it’s about people, process, and trust models working together.

⸻

🔗 Connect With Us

IT SPARC Cast

@ITSPARCCast on X

https://www.linkedin.com/company/sparc-sales/ on LinkedIn

John Barger

@john_Video on X

https://www.linkedin.com/in/johnbarger/ on LinkedIn

Lou Schmidt

@loudoggeek on X

https://www.linkedin.com/in/louis-schmidt-b102446/ on LinkedIn

Hosted on Acast. See acast.com/privacy for more information.

In this episode of IT SPARC Cast – Interview, John Barger sits down with Dr. Ewelina Kurtys of Final Spark to explore one of the most futuristic ideas in computing: building computers from living neurons.

Final Spark is a Swiss startup working to create biological computing systems using neurons derived from human stem cells. The goal is to develop a new form of compute that is dramatically more energy-efficient than traditional silicon—potentially by orders of magnitude.

In this conversation, John and Dr. Kurtys explore how neurons are sourced, how they are interfaced with traditional systems, and what it will take to build neuron-based data centers. They also discuss the challenges of programming biological systems, the timeline for commercialization, and what enterprise IT professionals should be doing today to prepare for this emerging paradigm.

This is a deep dive into the intersection of biology, AI, and infrastructure—and what could become the next major evolution of computing.  

⸻

⏱️ Show Notes

00:00 – Intro

An introduction to Final Spark and the concept of building computing systems using living neurons as an alternative to traditional silicon-based infrastructure.

⸻

❓ Questions

00:32 - Who Is Final Spark?

01:00 - How Do You Source Your Neurons?

01:43 - Neuron Quality Control

02:43 - Neurons In AI Data Centers

03:14 - Benefit Of Using Neurons

04:19 - When Will Neuron Based Compute Be Commercially Available

05:43 - Operating System Or Programming Language For Neurons

06:49 - What Does A Neuron Based Data Center Look Like?

07:55 - Containment And Security

08:28 - Data Persistence And Memory Erasure

09:10 - What Should IT Professionals Do Today To Prepare?

12:04 - How Does A Start-Up Get Involved Today?

12:44 - How Do You Program Neurons “Bits”? Are They Binary?

14:54 - How Do You Connect Neurons To Silicon Based Compute?

16:00 - Final Thoughts from Dr. Kurtys

⸻

https://www.finalspark.com

https://finalspark.com/articles/

⸻

🔁 Wrap Up

17:19 – Wrap Up

John reflects on the interview and the long-term implications of neuron-based computing. While still early-stage, the technology represents a potential shift in how compute is delivered—driven by energy efficiency, biological processing models, and new programming paradigms.

⸻

🔗 Connect With Us

IT SPARC Cast

@ITSPARCCast on X

https://www.linkedin.com/company/sparc-sales/ on LinkedIn

John Barger

@john_Video on X

https://www.linkedin.com/in/johnbarger/ on LinkedIn

Lou Schmidt

@loudoggeek on X

https://www.linkedin.com/in/louis-schmidt-b102446/ on LinkedIn

Hosted on Acast. See acast.com/privacy for more information.

Millions of Pokémon Go players unknowingly helped train real-world delivery robots using billions of images. Meanwhile, OpenAI’s ambitious Stargate data center expansion hits a major setback, highlighting the challenges of scaling AI infrastructure.

Nvidia pushes the frontier even further with plans for orbital AI data centers powered by its new Vera Rubin Space-1 chip system, while a growing movement suggests the “SaaS apocalypse” may be underway, driven by AI and open-source alternatives reshaping how software is built and consumed.

From crowdsourced AI training to space-based compute and the future of enterprise software, this episode explores where the next wave of IT disruption is coming from.  

Show Notes

00:00 – Intro

📰 News Bytes

00:45 – Pokémon Go Players Unknowingly Trained Delivery Robots With 30 Billion Images

Niantic has leveraged years of Pokémon Go gameplay data—over 30 billion images captured by users—to build a highly accurate Visual Positioning System (VPS) capable of centimeter-level location accuracy.

The discussion highlights both the brilliance of this crowdsourced data model and broader concerns around data ownership, enterprise data exposure, and unintended data usage.

https://www.popsci.com/technology/pokemon-go-delivery-robots-crowdsourcing/?utm_source=chatgpt.com 

⸻

07:18 – OpenAI’s Massive Stargate Data Center Expansion Canceled

Plans to expand a major AI data center tied to the Stargate initiative have been canceled, underscoring the complexity of building large-scale AI infrastructure.

Despite the cancellation, demand for AI compute remains extremely high, with other organizations potentially stepping in to utilize available capacity—reinforcing that AI infrastructure demand still far exceeds supply.

https://www.tomshardware.com/tech-industry/artificial-intelligence/openais-massive-stargate-data-center-canceled-as-firm-cant-reach-terms-with-oracle-operator-struggles-with-reliability-issues-meta-said-to-be-interested-in-snatching-excess-capacity 

⸻

11:06 – Nvidia Announces Vera Rubin Space-1 Chip System for Orbital AI Data Centers

Nvidia is pushing AI infrastructure beyond Earth with its Vera Rubin Space-1 system, designed for use in orbital data centers.

While challenges remain—especially around cooling and radiation—this represents a major step toward space-based AI infrastructure as demand for compute continues to surge.

https://www.cnbc.com/2026/03/16/nvidia-chips-orbital-data-centers-space-ai.html 

⸻

17:50 – The SaaS Apocalypse Is Open Source’s Greatest Opportunity

A growing trend suggests that traditional SaaS models may be under pressure as AI dramatically lowers the cost of building custom software.

The hosts highlight real-world examples of AI enabling individuals to build production-ready applications in hours, signaling a potential return to highly customized, in-house systems—powered by AI instead of large dev teams.

https://hackernoon.com/the-saas-apocalypse-is-opensources-greatest-opportunity 

⸻

🔁 Wrap Up

25:28 – Mail Bag

Listener Tim flags an issue with a previous episode upload, helping quickly resolve a distribution problem. A reminder of how valuable engaged listeners are to maintaining quality and consistency.

⸻

26:52 – Wrap Up

John and Lou close with thoughts on how rapidly the IT landscape is evolving—from AI-driven infrastructure and orbital compute to the reinvention of software delivery models—and encourage listeners to stay adaptable as these shifts accelerate.

⸻

🔗 Connect With Us

IT SPARC Cast

@ITSPARCCast on X

https://www.linkedin.com/company/sparc-sales/ on LinkedIn

John Barger

@john_Video on X

https://www.linkedin.com/in/johnbarger/ on LinkedIn

Lou Schmidt

@loudoggeek on X

https://www.linkedin.com/in/louis-schmidt-b102446/ on LinkedIn

Hosted on Acast. See acast.com/privacy for more information.

These actors are not just external attackers — they are getting hired, accessing corporate systems, and creating persistent insider threats that are extremely difficult to detect.

The episode explores how the scheme works, why traditional security controls fail, and what enterprise IT teams must do to defend against this evolving attack vector.

⸻

📝 Show Notes

A new cybersecurity threat is emerging that flips the traditional attack model on its head.

Instead of breaking into your network, attackers are getting hired into your company.

In this episode of IT SPARC Cast – CVE of the Week, John Barger and Lou Schmidt analyze the growing threat of North Korean IT worker schemes, where operatives pose as legitimate remote employees to gain direct access to enterprise systems.

⸻

🔎 How the Scheme Works

Threat actors:

•Apply for remote IT jobs using stolen or synthetic identities

•Pass interviews and onboarding processes

•Gain legitimate access to corporate systems

•Use that access to exfiltrate data, generate revenue, or stage future attacks

These individuals often work through:

•VPN masking

•Proxy networks

•Identity laundering through third parties

Once inside, they operate as trusted insiders, making detection significantly more difficult than traditional external threats.

⸻

⚠ Why This Is So Dangerous

This is not a vulnerability in software — it’s a failure in process, identity, and trust models.

Key risks include:

•Direct access to internal systems and data

•Ability to bypass perimeter security controls

•Long-term persistence without detection

•Potential for data exfiltration, espionage, or ransomware staging

Unlike typical breaches, these actors are:

•Authenticated

•Approved

•Operating under legitimate credentials

⸻

🏢 Enterprise IT Impact

This threat directly impacts:

•Remote-first organizations

•Companies hiring globally

•Teams using contractors or third-party staffing firms

•Organizations without strict identity verification processes

If your company hires remote developers, engineers, or IT staff — this is your problem.

⸻

🔐 Key Security Takeaways

To mitigate this risk, organizations should:

•Strengthen identity verification during hiring

•Require multi-factor authentication across all systems

•Monitor for unusual behavior from “trusted” accounts

•Implement least-privilege access controls

•Audit remote employee access regularly

•Coordinate with HR on security-aware hiring practices

This is a cross-functional problem — IT, Security, and HR must work together.

⸻

🔗 Source Article

https://www.nbcnews.com/investigations/north-korea-it-worker-scheme-nisos-fbi-rcna245025

⸻

🔗 Connect With Us

IT SPARC Cast

@ITSPARCCast on X

https://www.linkedin.com/company/sparc-sales/ on LinkedIn

John Barger

@john_Video on X

https://www.linkedin.com/in/johnbarger/ on LinkedIn

Lou Schmidt

@loudoggeek on X

https://www.linkedin.com/in/louis-schmidt-b102446/ on LinkedIn

Hosted on Acast. See acast.com/privacy for more information.

In this episode of IT SPARC Cast – News Bytes, John Barger & Lou Schmidt explore four major stories shaping enterprise IT, open-source software, AI infrastructure, and the future of data centers.

A new report reveals that two-thirds of Node.js installations are running outdated versions, creating major security and operational risks across modern software stacks. Meanwhile, Meta hires the creators of Moltbook, a platform designed as a social network for AI agents to communicate and collaborate.

Nvidia enters the agentic AI race with plans for an open-source enterprise AI agent platform, while Oracle’s massive investments in AI data centers spark debate about whether the industry is heading toward an infrastructure bubble.

From open-source sustainability to AI infrastructure strategy, this episode breaks down what these developments mean for enterprise IT leaders, developers, and technology investors. 

Show Notes

00:00 – Intro

📰 News Bytes

00:43 – Two Thirds of Node.js Installations Are Outdated

A new report from the OpenJS Foundation reveals that roughly two-thirds of Node.js deployments are running outdated or end-of-life versions, creating serious security and stability concerns across modern applications.

To address this, the Node.js LTS Upgrade and Modernization Program is connecting enterprises with trusted service providers audit, plan, and modernize their deployments.

The initiative also helps fund open-source development by directing a portion of service revenue back to the OpenJS Foundation.

https://openjsf.org/blog/nodejs-lts-upgrade-program

04:59 – Meta Hires the Duo Behind Moltbook

Meta has hired the creators of Moltbook, a platform designed as a collaboration network where AI agents can verify identity, exchange information, and coordinate tasks.

Meta’s move suggests a strategy to become the central hub for AI agent interaction, positioning the company to support a future where large numbers of autonomous software agents perform tasks for individuals and businesses.

https://www.axios.com/2026/03/10/meta-facebook-moltbook-agent-social-network

10:20 – Nvidia to Launch an Open-Source AI Agent Platform

Nvidia is preparing to release NemoClaw, an open-source AI agent platform designed to help enterprises deploy autonomous agents capable of automating workflows, managing data, and performing complex multi-step tasks.

Key aspects of the platform include:

• Enterprise-focused agent orchestration

• Open-source accessibility

• Compatibility beyond Nvidia hardware

• Integration with major enterprise software vendors

The move signals Nvidia’s growing interest in the agentic AI ecosystem, which could dramatically increase demand for GPU-accelerated compute infrastructure.

https://www.wired.com/story/nvidia-planning-ai-agent-platform-launch-open-source/

⸻

14:27 – Oracle Is Building Yesterday’s Data Centers With Tomorrow’s Debt

Oracle is investing heavily in new AI data centers, financing much of the expansion through debt as it competes with other hyperscale cloud providers.

Some analysts have raised concerns that rapid advances in AI hardware could outpace the construction timelines of new facilities, potentially creating financial risk.

However, the hosts point out that building data centers requires long lead times for power infrastructure, networking, and facilities, while the compute hardware itself is typically installed later in the deployment process.

The discussion highlights the importance of evaluating technology investment stories critically and considering both infrastructure realities and market narratives.

https://www.cnbc.com/2026/03/09/oracle-is-building-yesterdays-data-centers-with-tomorrows-debt.html

⸻

20:14 – Wrap up

⸻

🔗 Connect With Us

IT SPARC Cast

@ITSPARCCast on X

https://www.linkedin.com/company/sparc-sales/ on LinkedIn

John Barger

@john_Video on X

https://www.linkedin.com/in/johnbarger/ on LinkedIn

Lou Schmidt

@loudoggeek on X

https://www.linkedin.com/in/louis-schmidt-b102446/ on LinkedIn

Hosted on Acast. See acast.com/privacy for more information.

A new report reveals that two-thirds of Node.js installations are running outdated versions, creating major security and operational risks across modern software stacks. Meanwhile, Meta hires the creators of Moltbook, a platform designed as a social network for AI agents to communicate and collaborate.

Nvidia enters the agentic AI race with plans for an open-source enterprise AI agent platform, while Oracle’s massive investments in AI data centers spark debate about whether the industry is heading toward an infrastructure bubble.

From open-source sustainability to AI infrastructure strategy, this episode breaks down what these developments mean for enterprise IT leaders, developers, and technology investors.  

00:00 – Intro

📰 News Bytes

00:43 – Two Thirds of Node.js Installations Are Outdated

A new report from the OpenJS Foundation reveals that roughly two-thirds of Node.js deployments are running outdated or end-of-life versions, creating serious security and stability concerns across modern applications.

To address this, the Node.js LTS Upgrade and Modernization Program is connecting enterprises with trusted service providers that can:

• Audit existing deployments

• Plan phased upgrades

• Modernize dependencies

• Maintain production stability

https://openjsf.org/blog/nodejs-lts-upgrade-program

04:59 – Meta Hires the Duo Behind Moltbook

Meta has hired the creators of Moltbook, a platform designed as a collaboration network where AI agents can verify identity, exchange information, and coordinate tasks.

Meta’s move suggests a strategy to become the central hub for AI agent interaction, positioning the company to support a future where large numbers of autonomous software agents perform tasks for individuals and businesses.

https://www.axios.com/2026/03/10/meta-facebook-moltbook-agent-social-network

10:20 – Nvidia to Launch an Open-Source AI Agent Platform

Nvidia is preparing to release NemoClaw, an open-source AI agent platform designed to help enterprises deploy autonomous agents capable of automating workflows, managing data, and performing complex multi-step tasks.

The move signals Nvidia’s growing interest in the agentic AI ecosystem, which could dramatically increase demand for GPU-accelerated compute infrastructure.

https://www.wired.com/story/nvidia-planning-ai-agent-platform-launch-open-source/

14:27 – Oracle Is Building Yesterday’s Data Centers With Tomorrow’s Debt

Oracle is investing heavily in new AI data centers, financing much of the expansion through debt as it competes with other hyperscale cloud providers.

Some analysts have raised concerns that rapid advances in AI hardware could outpace the construction timelines of new facilities, potentially creating financial risk.

The discussion highlights the importance of evaluating technology investment stories critically and considering both infrastructure realities and market narratives.

https://www.cnbc.com/2026/03/09/oracle-is-building-yesterdays-data-centers-with-tomorrows-debt.html

🔁 Wrap Up

20:14 – Mail Bag

Listener Mel asks whether laser-based networking technologies, like the TaaraConnect system discussed in a previous episode, could help improve internet access in mountainous rural areas.

While line-of-sight laser connectivity could offer high speeds, weather conditions like fog and cloud cover could require backup connections such as radio or wired infrastructure.

⸻

🔗 Connect With Us

IT SPARC Cast

@ITSPARCCast on X

https://www.linkedin.com/company/sparc-sales/ on LinkedIn

John Barger

@john_Video on X

https://www.linkedin.com/in/johnbarger/ on LinkedIn

Lou Schmidt

@loudoggeek on X

https://www.linkedin.com/in/louis-schmidt-b102446/ on LinkedIn

Hosted on Acast. See acast.com/privacy for more information.

In this episode of IT SPARC Cast – CVE of the Week, John Barger and Lou Schmidt examine the KadNap router malware, which targets unpatched ASUS routers and installs a persistent backdoor designed to survive typical remediation efforts.

The malware was identified by researchers at Lumen’s Black Lotus Labs, who discovered that infected routers are being used as part of a botnet capable of proxying internet traffic and enabling other malicious activities.

Unlike many botnets that rely on centralized command servers, KadNap uses peer-to-peer control mechanisms similar to BitTorrent, making it significantly harder for security teams to disrupt.

⸻

🔎 What the KadNap Router Malware Does

The malware exploits vulnerabilities in ASUS routers that have not been patched or configured securely.

Once installed, KadNap:

•Creates a persistent backdoor on the router

•Survives reboots and firmware updates

•Enables remote control of the router

•Connects the device to a distributed botnet network

•Routes malicious traffic through compromised residential internet connections

Researchers also discovered the infected routers are being used by a fee-based proxy service called Doppelganger, allowing customers to route their internet traffic through unsuspecting victims’ home networks.

⸻

⚠ Why This Is Dangerous

Because the traffic originates from compromised home routers, victims could unknowingly appear responsible for malicious activity such as:

•Network attacks

•Surveillance operations

•Illegal browsing activity

•Staging points for additional cyber intrusions

This makes detection and attribution far more difficult.

⸻

🏢 Enterprise IT Risk

This vulnerability is not limited to home users.

ASUS also produces small-business routers, meaning organizations or small offices using these devices could be exposed.

IT professionals should also remember that compromised routers can provide attackers with a network foothold for lateral movement, especially if IoT or remote-user networks are poorly segmented.

⸻

🛠 How to Detect and Remove KadNap

Security experts recommend checking routers for signs of compromise:

Look for:

•SSH enabled unexpectedly

•Remote administration enabled

•Unknown certificates or scheduled tasks

•Suspicious entries in device logs

Because the malware attaches to configuration files, simply rebooting or restoring a configuration backup will not remove it.

The proper remediation process:

1.Perform a full factory reset

2.Update the router firmware immediately

3.Manually reconfigure the router (do not restore backups)

Experts also recommend changing default internal network ranges, such as moving away from the common 192.168.1.x subnet.

⸻

🔗 Source Article

https://arstechnica.com/security/2026/03/14000-routers-are-infected-by-malware-thats-highly-resistant-to-takedowns/

⸻

🔗 Connect With Us

IT SPARC Cast

@ITSPARCCast on X

https://www.linkedin.com/company/sparc-sales/ on LinkedIn

John Barger

@john_Video on X

https://www.linkedin.com/in/johnbarger/ on LinkedIn

Lou Schmidt

@loudoggeek on X

https://www.linkedin.com/in/louis-schmidt-b102446/ on LinkedIn

Hosted on Acast. See acast.com/privacy for more information.

Google begins testing a new quantum-resistant HTTPS certificate approach designed to defend the web against future quantum computing attacks. Meanwhile, Qualcomm’s CEO declares that the coming 6G mobile revolution will be essential for AI-driven applications, promising speeds up to 1 Tbps and sub-millisecond latency. Finally, TaaraConnect introduces a 25-Gbps laser-based networking system capable of delivering fiber-like speeds between buildings without laying cable.

From quantum-safe encryption to AI-driven wireless networks and laser communication links, this episode explores how the next generation of connectivity will reshape enterprise IT infrastructure, data centers, and global networks.  

⸻

Show Notes

00:00 – Intro

⸻

📰 News Bytes

00:51 – Google Quantum-Proofs HTTPS with Compact Certificates

Google and Cloudflare are testing a new method to make HTTPS certificates resistant to future quantum attacks.

Instead of traditional signature chains, the system uses Merkle Tree Certificates (MTCs) to dramatically shrink quantum-safe cryptographic data from roughly 15 KB down to about 700 bytes, making it practical for real-world internet use.

The experiment begins with about 1,000 TLS certificates, with standards work underway through the IETF. If successful, this approach could become a foundational component of post-quantum internet security.

The big unknown: how much processing overhead these new cryptographic methods will require on older client devices.

https://arstechnica.com/security/2026/02/google-is-using-clever-math-to-quantum-proof-https-certificates/

⸻

05:42 – Qualcomm CEO Says the 6G Revolution Is Coming

At Mobile World Congress, Qualcomm CEO Cristiano Amon argued that the AI era will demand 6G networks, dramatically increasing bandwidth and lowering latency.

Projected 6G capabilities include:

• Speeds of 100 Gbps to 1 Tbps

• Sub-millisecond latency

• Massive connectivity for billions of devices

• AI-driven network management

The shift toward AI-heavy cloud processing means network traffic will become even more north-south oriented, sending massive datasets between edge devices and cloud infrastructure.

Early 6G trials are expected around 2028, with broader deployments beginning around 2029.

https://fortune.com/2026/03/03/qualcomm-ceo-resistance-is-futile-6g-mobile-revolution-approaches/

⸻

13:31 – TaaraConnect Uses Lasers to Deliver 25-Gbps Internet

TaaraConnect is developing a laser-based point-to-point networking system capable of delivering up to 25 Gbps over distances of about 6 miles (10 km).

Instead of radio waves, the system transmits data using focused beams of light, creating fiber-like connectivity without physical cables.

Key features include:

• Adaptive beam alignment

• Automatic power adjustments during interference

• Redundant beam paths to mitigate disruptions

• Low-latency high-bandwidth links between buildings

While heavy fog remains the primary limitation, the technology could provide a powerful alternative to expensive metro fiber deployments, particularly in dense urban environments.

https://newatlas.com/telecommunications/google-taara-25gbps-internet-cities-light/

⸻

🔁 Wrap Up

19:31 – Mail Bag

Listener Xavier shares thoughts on the rising cost of compute power and suggests that bio-AI systems using living neurons could eventually deliver superior performance-per-watt compared to traditional silicon chips.

The discussion highlights a growing industry focus on compute efficiency and power consumption as data center infrastructure scales to support AI workloads.

⸻

🔗 Connect With Us

IT SPARC Cast

@ITSPARCCast on X

https://www.linkedin.com/company/sparc-sales/ on LinkedIn

John Barger

@john_Video on X

https://www.linkedin.com/in/johnbarger/ on LinkedIn

Lou Schmidt

@loudoggeek on X

https://www.linkedin.com/in/louis-schmidt-b102446/ on LinkedIn

Hosted on Acast. See acast.com/privacy for more information.

The issue centers around CVE-2026-21385, a high-severity vulnerability affecting Qualcomm graphics components used in Android devices. While the vulnerability requires physical access, it is actively being exploited in the wild, making it a serious concern for enterprise IT environments.

But the real story isn’t smartphones.

The bigger risk lies in Android devices hiding in plain sight across enterprise infrastructure — including point-of-sale terminals, warehouse scanners, embedded industrial systems, and other IoT devices that often run outdated Android versions and rarely receive timely security updates.

⸻

🔎 CVE-2026-21385 Overview

•CVE: CVE-2026-21385

•Severity: High (CVSS 7.8)

•Component: Qualcomm GPU graphics driver used in Android

•Exploit Status: Actively exploited in the wild

•Access Required: Physical access

•Patch: Included in March 2026 Android Security Bulletin

Several additional vulnerabilities were also patched in the same release, including critical Android framework remote code execution flaws, increasing the urgency for organizations to deploy updates wherever possible.

⸻

⚠ Why Enterprise IT Should Care

Most organizations focus on employee phones when thinking about Android security.

However, the real exposure often comes from embedded Android devices that organizations forget about:

Common examples include:

•Point-of-sale payment terminals

•Warehouse inventory scanners (Zebra, Honeywell, etc.)

•Retail handheld devices

•Industrial control panels

•Vehicle infotainment systems running Android

•Embedded tablets in appliances or machinery

Many of these devices:

•Run older Android versions

•Receive delayed or nonexistent updates

•Expose USB or physical ports that could enable exploitation

•Are connected to internal networks

If compromised, these systems could become the first step in a lateral network attack.

⸻

🔐 Key Security Takeaways

Organizations should treat this vulnerability as a wake-up call for Android-based IoT security.

Recommended actions:

•Inventory all Android-based devices in your environment

•Identify IoT or embedded Android systems

•Verify whether vendors provide security updates

•Push vendors for timelines if patches are not available

•Segregate IoT devices onto isolated networks

•Lock down physical access and exposed USB ports

Ignoring embedded Android devices can create a hidden attack path directly into corporate networks.

⸻

💬 Listener Feedback

Following last week’s episode discussing the Conduent ransomware breach, listeners shared their experiences receiving breach notification letters.

One listener reported receiving a notification despite not participating in government assistance programs, while another reported being impacted through health insurance providers like Blue Cross Blue Shield.

The scope of the Conduent breach appears to be continuing to expand, reinforcing the importance of monitoring vendor supply-chain exposure.

⸻

🔗 Connect With Us

IT SPARC Cast

@ITSPARCCast on X

https://www.linkedin.com/company/sparc-sales/ on LinkedIn

John Barger

@john_Video on X

https://www.linkedin.com/in/johnbarger/ on LinkedIn

Lou Schmidt

@loudoggeek on X

https://www.linkedin.com/in/louis-schmidt-b102446/ on LinkedIn

Hosted on Acast. See acast.com/privacy for more information.

Anthropic revises its flagship AI safety pledge amid competitive pressure, open source registries warn they lack funding for basic security protections, and OpenAI scrambles for compute power as large-scale infrastructure plans stall.

From AI governance and supply chain risk to infrastructure bottlenecks and power constraints, this episode explores what enterprise IT leaders need to be watching right now.

⸻

⏱️ Show Notes

00:00 – Intro

Anthropic revises its safety framework, open source ecosystems struggle to fund security, and OpenAI races to secure compute capacity as infrastructure constraints tighten across the AI industry.

⸻

📰 News Bytes

⸻

00:44 – Anthropic Drops Flagship Safety Pledge

Anthropic has revised its 2023 Responsible Scaling Policy, removing its categorical commitment to halt training if safety guarantees could not be ensured in advance. The company says the shift reflects rapid AI advancement, competitive pressures, and the need for transparency over unilateral restrictions.

John and Lou unpack what this means for enterprise deployments: Is this a rollback of safety? Or a move toward operational flexibility and published risk roadmaps? The bigger issue may be how AI vendors balance guardrails, customer control, and competitive pressure.

https://time.com/7380854/exclusive-anthropic-drops-flagship-safety-pledge/ 

⸻

08:16 – Open Source Registries Can’t Afford Basic Security

Major open source ecosystems such as PyPI, npm, RubyGems, and others are reportedly facing funding shortfalls that threaten their ability to implement fundamental security protections.

With supply chain attacks on the rise and AI accelerating code generation, underfunded registries present a growing enterprise risk. The hosts discuss why “free” does not mean costless — and why corporate IT teams must contribute financially or through engineering resources to sustain the security of the tools they depend on.

https://www.theregister.com/2026/02/16/open_source_registries_fund_security/ 

⸻

12:36 – Inside OpenAI’s Scramble to Secure Compute After Stargate Stalled

OpenAI’s large-scale infrastructure plans have reportedly slowed, forcing the company to seek alternative compute sources to sustain AI growth.

The episode explores the real bottlenecks: wafer starts, power generation, turbines, construction capacity, data center labor, and capital coordination. With AI revenue tightly correlated to compute availability, infrastructure constraints may be the biggest story in AI for 2026 and beyond.

https://www.theinformation.com/articles/inside-openais-scramble-get-computing-power-stargate-stalled 

⸻

🔁 Wrap Up

19:15 – Mail Bag

Listener Xavier highlights how surface-level headlines often hide deeper enterprise implications — a reminder that IT leaders must look beneath the story to understand downstream risk and opportunity.

20:36 – Wrap Up

From AI safety governance and competitive pressure to supply chain funding gaps and compute shortages, Episode 27 reinforces one theme: infrastructure, transparency, and long-term planning now define enterprise AI strategy.

⸻

🔗 Connect With Us

IT SPARC Cast

@ITSPARCCast on X

https://www.linkedin.com/company/sparc-sales/ on LinkedIn

John Barger

@john_Video on X

https://www.linkedin.com/in/johnbarger/ on LinkedIn

Lou Schmidt

@loudoggeek on X

https://www.linkedin.com/in/louis-schmidt-b102446/ on LinkedIn

Hosted on Acast. See acast.com/privacy for more information.

What began as reports of a 10.5 million record breach has now escalated to an estimated 25 million impacted individuals, with the ransomware group SafePay claiming responsibility and alleging over 8 terabytes of data exfiltrated.

⸻

🔎 What Happened?

Conduent, which provides backend processing for government assistance programs and health benefits, confirmed that sensitive personal and corporate information may have been exposed.

Reported exposed data includes:

•Names

•Dates of birth

•Addresses

•Social Security numbers

•Employment records

•Financial information

•Medical and health insurance details

•Internal business documents

SafePay ransomware actors reportedly gained access through compromised credentials and then moved laterally through Conduent’s systems.

This is a textbook example of a chained cyberattack, where one small compromise enables full-scale enterprise exposure.

⸻

🌎 Scope of the Impact

The breach affects multiple U.S. states and programs, including:

•Texas (~15.4 million impacted)

•Oregon (~10.5 million impacted)

•Delaware

•Massachusetts

•New Hampshire

•Georgia

•South Carolina

•New Jersey

•Maine

•New Mexico

Programs potentially affected:

•Medicaid

•SNAP / EBT food assistance

•Unemployment benefits

•Health insurance processing (including Blue Cross Blue Shield and Humana)

•Corporate employee benefit programs

Additionally, approximately 17,000 Volvo Group North America employees may have been impacted.

⸻

⚠ Why This Matters for Enterprise IT

This is not “just” a public-sector breach.

Many private companies rely on Conduent for backend benefits processing. If your organization uses:

•Blue Cross Blue Shield

•Humana

•Third-party HR / benefits processors

You must immediately:

•Contact your HR and benefits teams

•Request incident briefings from vendors

•Determine if employee data was exposed

•Prepare remediation and communication plans

⸻

🔐 Security Lessons

•Credential compromise remains a primary entry point

•Lateral movement amplifies initial footholds

•Ransomware groups continue combining encryption with large-scale data exfiltration

•Transparency and timely disclosure are critical

Conduent acknowledged the breach, engaged forensic investigators, and notified impacted parties — a necessary and responsible response.

⸻

💬 Listener Feedback

The episode also includes feedback from Kevin regarding last week’s Apple iOS 26 patch discussion. While some users hesitate to upgrade due to UI and stability concerns, security patches addressing critical vulnerabilities must take priority.

⸻

🔗 Connect With Us

IT SPARC Cast

@ITSPARCCast on X

https://www.linkedin.com/company/sparc-sales/ on LinkedIn

John Barger

@john_Video on X

https://www.linkedin.com/in/johnbarger/ on LinkedIn

Lou Schmidt

@loudoggeek on X

https://www.linkedin.com/in/louis-schmidt-b102446/ on LinkedIn

Hosted on Acast. See acast.com/privacy for more information.

Anthropic and the Pentagon clash over Claude’s military use, Western Digital reports zero remaining HDD capacity for 2026, and Microsoft confirms a Copilot bug that summarized confidential emails. From supply chain strategy to SaaS risk management, this episode highlights why enterprise IT leaders must think beyond features and focus on contracts, capacity, and control.

⸻

⏱️ Show Notes

00:00 – Intro

Hard drive shortages, AI contract battles, and Copilot privacy concerns headline a week that reinforces one theme: control over infrastructure and software matters more than ever.

⸻

📰 News Bytes

00:46 – Anthropic and the Pentagon Are Reportedly Arguing Over Claude Usage

Anthropic pushes back against unrestricted military use of Claude AI, raising ethical, contractual, and operational questions. The Pentagon may reconsider its $200M relationship, exposing a major risk for organizations deploying AI: what happens when vendor policies change after integration?

https://techcrunch.com/2026/02/15/anthropic-and-the-pentagon-are-reportedly-arguing-over-claude-usage/ 

⸻

07:19 – Western Digital Has No More HDD Capacity Left for 2026

Western Digital reports its entire 2026 hard drive production is already spoken for. Similar signals from Seagate suggest storage pricing pressure is imminent. The hosts explain why this isn’t just about spinning disks—it’s about AI data center demand driving up costs across RAM, SSDs, GPUs, and enterprise hardware.

https://wccftech.com/western-digital-has-no-more-hdd-capacity-left-out/ 

⸻

12:06 – Microsoft Says Bug Causes Copilot to Summarize Confidential Emails

Microsoft confirms a Copilot bug that processed confidential emails stored in drafts and sent folders, despite policy settings meant to block them. Although no data reportedly left the organization, the incident underscores governance, SaaS dependency, and AI access-control risks enterprises must plan for.

https://www.bleepingcomputer.com/news/microsoft/microsoft-says-bug-causes-copilot-to-summarize-confidential-emails/ 

⸻

🔁 Wrap Up

16:42 – Mail Bag

Listener Dennis drops a Back to the Future “jigawatt” reference, and Xavier reinforces the importance of AI security hygiene and fine-grained permission management.

17:52 – Wrap Up

Final thoughts on vendor lock-in, AI policy control, supply chain modeling, and why IT leaders need stronger collaboration with finance and legal teams.

⸻

🔗 Connect With Us

IT SPARC Cast

@ITSPARCCast on X

https://www.linkedin.com/company/sparc-sales/

John Barger

@john_Video on X

https://www.linkedin.com/in/johnbarger/

Lou Schmidt

@loudoggeek on X

https://www.linkedin.com/in/louis-schmidt-b102446/

Hosted on Acast. See acast.com/privacy for more information.

The focus: CVE-2026-20700, a memory corruption flaw in Apple’s dynamic link layer that could allow attackers to break out of the sandbox and achieve remote code execution (RCE).

Although exploitation requires physical access, the definition of “physical” in today’s hybrid enterprise world is broader than it sounds. Remote management tools, compromised accounts, lost devices, or improperly secured BYOD endpoints can all create real-world exposure.

With Apple’s unified “26” operating system line now spanning every platform, this patch affects:

•iOS 26.3

•iPadOS 26.3

•macOS 26.3

•watchOS 26.3

•tvOS 26.3

•visionOS 26.3

Security researchers are classifying this vulnerability as critical/high severity, and enterprises are urged to patch immediately.

⸻

🔎 CVE-2026-20700 Details

•Type: Memory corruption

•Impact: Sandbox escape → Remote Code Execution

•Exploit Path: Physical or logical device access

•Risk Level: High/Critical (no official CVSS published)

•Fix: Upgrade to Apple OS version 26.3

⸻

⚠ Why This Matters for Enterprise IT

1️⃣ BYOD Risk Surface

Bring-Your-Own-Device policies mean iPhones, iPads, and Macs often connect to corporate networks without full administrative control. A vulnerable device on your network increases lateral movement risk.

2️⃣ Physical Access Isn’t Just “Someone in the Room”

Remote tools, compromised Apple IDs, or stolen devices expand the meaning of physical access.

3️⃣ Upgrade Hesitation Is Real

Apple’s 26 release introduced major UI changes (including the controversial glass interface). Stability concerns have led some users to delay upgrades — increasing exposure time.

Security must outweigh aesthetic or usability concerns.

⸻

🛠 Enterprise Recommendations

•Immediately communicate required upgrade to 26.3

•Enforce OS minimum versions where possible

•Review BYOD policies and mobile device controls

•Audit Apple device access on corporate networks

•Educate users about lost/stolen device risk

⸻

💬 Listener Feedback

The episode also includes commentary from Chris, a general counsel and chief risk officer, who responded to last week’s Notepad RCE discussion. He raises an important point about expanding application functionality increasing attack surface — a lesson that applies here as well.

⸻

🔗 Connect With Us

IT SPARC Cast

@ITSPARCCast on X

https://www.linkedin.com/company/sparc-sales/ on LinkedIn

John Barger

@john_Video on X

https://www.linkedin.com/in/johnbarger/ on LinkedIn

Lou Schmidt

@loudoggeek on X

https://www.linkedin.com/in/louis-schmidt-b102446/ on LinkedIn

Hosted on Acast. See acast.com/privacy for more information.

Then things escalate: hyperscalers begin signing real contracts for next-generation nuclear power, and Elon Musk gets serious about orbital data centers—suggesting that the future of compute may extend beyond the planet. If you’re tracking AI infrastructure, network evolution, and the power constraints shaping the industry, this episode connects the dots.

⸻

⏱️ Show Notes

00:00 – Intro

John and Lou preview a week dominated by UniFi upgrades, federal energy discussions, nuclear power tipping points, and serious momentum toward data centers in space.

⸻

📰 News Bytes

00:44 – UniFi Network 10.1

Ubiquiti releases UniFi Network 10.1 with major stability and scalability improvements, Wi-Fi Doctor diagnostics, UI refinements, enhanced policy visibility, and optimizations for Wi-Fi 7 and multi-gig deployments. The hosts discuss why UniFi continues its march toward true enterprise credibility while remaining accessible for SMB and prosumer environments.

https://blog.ui.com/article/introducing-unifi-network-10-1 

⸻

05:13 – White House Eyes Data Center Agreements Amid Energy Price Spikes

As AI data center expansion drives regional energy price pressure, the White House explores voluntary agreements with major tech companies to shift infrastructure costs away from consumers. The conversation explores the economics of AI growth, the inevitability of nuclear power, and whether energy becomes the defining constraint of the AI race.

https://www.politico.com/news/2026/02/09/trump-administration-eyes-data-center-agreements-amid-energy-price-spikes-00772024 

⸻

09:02 – Next-Gen Nuclear’s Tipping Point: Meta and Hyperscalers Sign Deals

Meta and other hyperscalers begin signing legally binding agreements with next-generation nuclear companies like TerraPower and Oklo. John and Lou explain why signed contracts—not press releases—mark the true tipping point for small modular reactors powering AI infrastructure.

https://www.aol.com/articles/next-gen-nuclear-tipping-point-214209248.html 

⸻

11:34 – Elon Musk Gets Serious About Orbital Data Centers

Following strategic moves linking xAI and SpaceX, Musk pivots attention toward orbital and lunar infrastructure. The hosts unpack the logic behind space-based data centers, cooling challenges, Starlink integration, and why the economics may be less crazy than they first appear.

https://techcrunch.com/2026/02/05/elon-musk-is-getting-serious-about-orbital-data-centers/ 

⸻

🔁 Wrap Up

18:59 – Mail Bag

Listener Jonah questions whether massive AI infrastructure financing signals a bubble. John and Lou explain why AI demand is currently compute-constrained—not hype-driven—and why any financial correction would look very different from the dot-com era.

21:52 – Wrap Up

Final thoughts on nuclear inevitability, orbital infrastructure, and the reality that energy—not chips—may define the next decade of enterprise IT.

⸻

🔗 Connect With Us

IT SPARC Cast

@ITSPARCCast on X

https://www.linkedin.com/company/sparc-sales/

John Barger

@john_Video on X

https://www.linkedin.com/in/johnbarger/

Lou Schmidt

@loudoggeek on X

https://www.linkedin.com/in/louis-schmidt-b102446/

Hosted on Acast. See acast.com/privacy for more information.

Yes — even Notepad isn’t safe anymore.

This vulnerability stems from a command injection flaw in the modern Windows 11 Store version of Notepad (11.x prior to patch). The issue allows malicious .md (Markdown) files containing crafted links or interactive content to execute arbitrary code when opened and clicked by a user.

With a CVSS score of 8.8, this vulnerability becomes especially dangerous when chained with other exploits.

⸻

🔎 What You Need to Know

CVE-2026-2841 – Windows Notepad RCE

•Affects: Windows 11 modern Notepad (Microsoft Store version 11.x prior to Patch Tuesday update)

•Does NOT affect: Legacy Notepad on Windows 10, Windows 7, or classic versions

•Attack Vector: Malicious .md file delivered via phishing

•Trigger: User opens file and clicks embedded link

•Impact: Remote Code Execution with user-level permissions

•Severity: CVSS 8.8 (High)

⸻

⚠ Why This Matters

•Perfect phishing vehicle: malicious Markdown attachment

•Executes arbitrary code under the user’s permissions

•Ideal for lateral movement in enterprise environments

•Dangerous when combined with other exploits

•Many organizations delay Patch Tuesday updates — this one should NOT wait

⸻

🛠 Mitigation & Recommendations

•Immediately update Notepad via Microsoft Store

•Audit Windows 11 endpoints for modern Notepad version

•Train users to avoid opening unknown .md attachments

•Consider simpler text editors for baseline editing tasks

•Evaluate enterprise endpoint protection against command injection vectors

⸻

💻 Alternative Editors (With Security Awareness)

John and Lou discuss safer editing alternatives including:

•Notepad++

•Visual Studio Code / Codeium

•Sublime Text

•Atom

•Vim / NeoVim / Emacs

•JetBrains IDEs

Reminder: More features = more attack surface.

⸻

💬 Wrap Up

John and Lou also respond to listener feedback from Andrew regarding their recent OpenClaw security discussion. They clarify their stance:

•They are not anti-AI.

•They are pro-security.

•Bleeding-edge tech requires controlled rollout and sandboxing.

•Enterprises must protect privileged data access.

Security-first thinking is not fear — it’s responsible IT leadership.

⸻

🔗 Connect With Us

IT SPARC Cast

@ITSPARCCast on X

https://www.linkedin.com/company/sparc-sales/ on LinkedIn

John Barger

@john_Video on X

https://www.linkedin.com/in/johnbarger/ on LinkedIn

Lou Schmidt

@loudoggeek on X

https://www.linkedin.com/in/louis-schmidt-b102446/ on LinkedIn

Hosted on Acast. See acast.com/privacy for more information.

They also break down why the AI boom is starting to squeeze Apple’s famously strong profit margins—and what that means for device pricing, silicon supply, and enterprise IT budgets. If you care about where compute, power, and economics collide, this episode connects the dots.

📌 Show Notes

00:00 – Intro

John and Lou kick off the episode with a look at why quantum cooling breakthroughs, massive AI cloud investments, and chip supply pressure on Apple all point to accelerating change across enterprise IT.

⸻

📰 News Bytes

00:55 – Chalmers University Makes Quantum Cool

Researchers at Chalmers University of Technology unveil a breakthrough quantum refrigeration method that uses controlled noise to improve cooling near absolute zero. The discussion explores why advances like this could rapidly reduce the cost and complexity of quantum computing and push it closer to real enterprise use cases.

https://www.sciencedaily.com/releases/2026/01/260129080418.htm 

⸻

06:10 – Oracle Raising Up to $50B for AI Cloud

Oracle plans to raise up to $50 billion to expand AI-focused cloud data centers as part of Project Stargate. John and Lou unpack why Oracle’s existing enterprise relationships give it a unique advantage—and why power and compute, not demand, may become the real limiting factors for AI growth.

https://www.techrepublic.com/article/news-oracle-50b-ai-cloud/ 

⸻

10:28 – The AI Boom Is Coming for Apple’s Profit Margins

The surge in AI-driven chip demand is putting pressure on Apple’s historically strong margins. As TSMC capacity is increasingly consumed by Nvidia, OpenAI, and hyperscalers, the hosts break down why Apple may face higher silicon costs—and what that means for device pricing, IT refresh cycles, and enterprise procurement.

https://www.msn.com/en-us/money/technology/the-ai-boom-is-coming-for-apple-s-profit-margins/ar-AA1VpgpA 

⸻

🔁 Wrap Up

17:11 – Mail Bag

Listener feedback sparks a nuanced discussion on hybrid work, mandatory office policies, and why management capability—not location—is often the real issue.

21:24 – Wrap Up

Final thoughts on quantum acceleration, AI infrastructure economics, and why IT leaders need to prepare for rising hardware costs and longer planning horizons.

⸻

🔗 Connect With Us

IT SPARC Cast

@ITSPARCCast on X

https://www.linkedin.com/company/sparc-sales/

John Barger

@john_Video on X

https://www.linkedin.com/in/johnbarger/

Lou Schmidt

@loudoggeek on X

https://www.linkedin.com/in/louis-schmidt-b102446/

Hosted on Acast. See acast.com/privacy for more information.

In this episode of IT SPARC Cast – CVE of the Week, John and Lou break down why running OpenClaw (and related platforms like MoltBook) on corporate hardware or with access to enterprise data is dangerous right now, even if the long-term vision is compelling.

The discussion centers on three types of OpenClaw users:

1.Sandbox Experimenters – Users running OpenClaw in isolated labs or test environments with no access to corporate data.

2.Dedicated VM / Hardware Users – Users running OpenClaw separately, but still granting it access to cloud services, email, or internal APIs.

3.Daily Driver Users – Users installing OpenClaw directly on work PCs and giving it full access to files, email, chat, and automation tools.

John and Lou argue that only the first group is safe today.

Groups #2 and #3 dramatically expand the attack surface, introducing risks such as credential exfiltration, indirect prompt injection, data leakage, and supply-chain style compromises via third-party “skills.”

The episode uses a “bio hotcell” analogy: OpenClaw can be used safely only when isolated, constrained, monitored, and treated as potentially hazardous. Without those controls, it becomes a silent data-exfiltration engine operating entirely inside allowed enterprise workflows.

The takeaway for IT leaders is clear:

HR and IT must act together now to define policies that prohibit OpenClaw and MoltBook from running on corporate devices or accessing corporate data until proper governance, tooling, and security controls exist.

⸻

🔚 Wrap Up & Links

Follow and connect with us:

IT SPARC Cast

@ITSPARCCast on X

https://www.linkedin.com/company/sparc-sales/ on LinkedIn

John Barger

@john_Video on X

https://www.linkedin.com/in/johnbarger/ on LinkedIn

Lou Schmidt

@loudoggeek on X

https://www.linkedin.com/in/louis-schmidt-b102446/ on LinkedIn

Hosted on Acast. See acast.com/privacy for more information.

The discussion connects culture, power, cooling, and compute—showing why AI growth isn’t just about models and GPUs, but about solving the physical and human constraints that come with them. If you’re responsible for enterprise IT strategy, infrastructure planning, or talent retention, this episode delivers context you won’t get from headlines alone.

⸻

⏱️ Show Notes

00:00 – Intro

John and Lou preview a packed episode covering remote-work backlash, radical new data-center cooling approaches, and Microsoft’s latest move to control its AI destiny with custom silicon.

⸻

📰 News Bytes

01:00 – Work-From-Office Mandate? Expect Top Talent Turnover and Culture Rot

New research highlighted by CIO Magazine shows that strict return-to-office mandates are driving increased attrition among top performers, longer hiring cycles, and declining trust. John and Lou unpack why “butts-in-seats” metrics fail modern organizations and how poor remote-management skills—not productivity—are often the real problem.

https://www.cio.com/article/4119562/work-from-office-mandate-expect-top-talent-turnover-culture-rot.html 

⸻

08:14 – L.A. Startup Uses SpaceX Tech to Cool Data Centers With Less Power and No Water

An LA-based startup is applying SpaceX rocket turbopump technology and supercritical CO₂ to dramatically reduce data-center cooling power, footprint, and water usage. The hosts explain why cooling—not chips—is becoming one of the biggest bottlenecks in AI expansion and how innovations like this could unlock sustainable growth.

https://finance.yahoo.com/news/l-startup-uses-spacex-tech-175628363.html

⸻

14:11 – Microsoft Announces a Powerful New Chip for AI Inference

Microsoft unveils the Maia 200, a custom AI inference accelerator built on TSMC’s 3-nm process with 100 billion transistors. John and Lou break down why inference-optimized chips matter, how this fits into a broader trend of hyperscalers building custom silicon, and why efficiency per watt is becoming the defining metric for AI at scale.

https://techcrunch.com/2026/01/26/microsoft-announces-powerful-new-chip-for-ai-inference/

⸻

🔁 Wrap Up

19:49 – Mail Bag

Listener feedback revisits classic operating systems, early AI roots, and why distributed computing concepts from decades ago are suddenly relevant again.

22:47 – Wrap Up

John and Lou close by emphasizing that AI’s future depends on solving power, cooling, and organizational challenges—not just shipping faster chips.

⸻

🔗 Connect With Us

IT SPARC Cast

@ITSPARCCast on X

https://www.linkedin.com/company/sparc-sales/ on LinkedIn

John Barger

@john_Video on X

https://www.linkedin.com/in/johnbarger/

Lou Schmidt

@loudoggeek on X

https://www.linkedin.com/in/louis-schmidt-b102446/ on LinkedIn

Hosted on Acast. See acast.com/privacy for more information.

First, we dive into a new lawsuit alleging that Meta can access or infer WhatsApp message contents, despite years of public claims that WhatsApp is fully end-to-end encrypted. We unpack what “access” really means in modern encrypted messaging systems, including metadata, client-side processing, backups, and enterprise risk implications—especially for organizations using WhatsApp for daily business communications.

https://www.bloomberg.com/news/articles/2026-01-25/lawsuit-claims-meta-can-see-whatsapp-chats-in-breach-of-privacy

Next, we examine a major data exposure involving Chat & Ask AI, a popular AI chatbot aggregator with tens of millions of users. Due to a backend Firebase misconfiguration, hundreds of millions of private conversations—including highly sensitive topics—were left publicly accessible. This incident highlights the growing risk of Shadow AI inside enterprises and the dangers of third-party AI wrappers that lack enterprise-grade security controls.

https://www.404media.co/massive-ai-chat-app-leaked-millions-of-users-private-conversations/

The episode closes with listener feedback on a previously covered UniFi Access vulnerability and a broader discussion on how organizations should educate, monitor, and protect users without resorting to blunt enforcement.

Hosted on Acast. See acast.com/privacy for more information.

The conversation connects enterprise software evolution, network architecture at scale, and the hard economic realities of AI infrastructure—especially power and compute. If you’re responsible for enterprise platforms, networking strategy, or long-term IT planning, this episode provides context that goes beyond the headlines.

⸻

⏱️ Show Notes

00:00 – Intro

John and Lou preview the episode, touching on ERP’s looming transformation, UniFi’s new Fabric approach, and why AI demand—especially at OpenAI—is driven by hard infrastructure realities, not hype.

⸻

📰 News Bytes

00:48 – ERP Isn’t Dead Yet – But Most Execs Are Planning the Wake

A survey of more than 4,300 executives shows growing skepticism about ERP’s long-term dominance, even as most organizations remain satisfied with current systems. John and Lou explain why AI-driven, modular, and agentic ERP models are likely evolutions—not rip-and-replace events—and what enterprise IT teams should be doing now to prepare.

https://www.theregister.com/2026/01/19/erp_survey_rimini_street/ 

⸻

06:28 – Ubiquiti Introduces UniFi Fabric

Ubiquiti unveils UniFi Fabric, a centralized orchestration layer designed to manage policies, identity-based networking, Zero Trust, and multi-site environments without cloud licensing. The discussion compares UniFi’s approach to traditional network fabrics like VXLAN and SPBM, highlighting why this controller-first model could appeal to MSPs and mid-sized enterprises.

https://blog.ui.com/article/introducing-unifi-fabrics 

⸻

14:14 – AI Bubble? Maybe. OpenAI Risk? Not Anytime Soon.

John breaks down why OpenAI’s revenue growth is directly tied to available compute capacity, not speculative demand. Using concrete megawatt, gigawatt, and ARR figures, the hosts explain why AI may see valuation corrections—but why companies like OpenAI, NVIDIA, and Anthropic are unlikely to disappear.

https://openai.com/index/a-business-that-scales-with-the-value-of-intelligence/ 

⸻

🔁 Wrap Up

24:16 – Mail Bag

Listener feedback reinforces the growing link between AI growth and power infrastructure, with discussion around electrical safety, regulation, and why energy expertise may be one of the most valuable skills in the coming decade.

27:39 – Wrap Up

John and Lou close with a reminder that enterprise IT leaders will increasingly be asked to validate power, nuclear, and infrastructure decisions at the executive level—and that staying informed now is critical.

⸻

🔗 Connect With Us

IT SPARC Cast

@ITSPARCCast on X

https://www.linkedin.com/company/sparc-sales/ on LinkedIn

John Barger

@john_Video on X

https://www.linkedin.com/in/johnbarger/

Lou Schmidt

@loudoggeek on X

https://www.linkedin.com/in/louis-schmidt-b102446/ on LinkedIn

Hosted on Acast. See acast.com/privacy for more information.

Rather than focusing on one vulnerability, VoidLink is designed to chain together many smaller flaws across Linux, containers, and cloud platforms like AWS, Azure, GCP, Docker, and Kubernetes—creating a stealthy, long-term access platform. Researchers believe VoidLink was developed rapidly using AI assistants, offering a rare look at how next-generation malware may be authored, iterated, and deployed. This episode explains why VoidLink matters, how defenders should think about chained exploits, and why this may be an early warning sign for the future of cloud and container security.

⸻

Show Notes (Podcast)

Episode Overview

This week’s CVE of the Week focuses on VoidLink, a newly identified Linux malware framework designed for persistence, stealth, and modular exploitation across cloud and container environments. While not a single CVE, VoidLink highlights how attackers are moving toward framework-driven, AI-assisted exploit chaining rather than isolated vulnerabilities.

Key Topics Covered

•What VoidLink is and why it’s different from traditional malware

•How chaining low-severity vulnerabilities can result in full compromise

•Targeted environments: Linux, Docker, Kubernetes, AWS, Azure, and GCP

•Use of loaders, implants, evasion techniques, and modular plugins

•Evidence suggesting AI-assisted development with rapid iteration

•Why this gives defenders a rare opportunity to observe a threat early in its lifecycle

•The implications for cloud security, container hardening, and future CVEs

Why This Matters

VoidLink represents a shift from one-off exploits to malware platforms—essentially an “IDE for hacking.” Understanding how these frameworks are built and how they operate is critical for anticipating future attacks and improving detection strategies before they become widespread.

⸻

Listener Feedback Highlight

We’d like to give a shout-out to Nihal for his thoughtful LinkedIn comment on our earlier Top 10 Operating System Failures episode—specifically his hot take defending Windows ME and critiquing Windows XP’s compatibility break. We love informed debate like this and appreciate listeners who challenge conventional wisdom.

⸻

Wrap-Up & Social Links

That wraps up this episode of IT SPARC Cast – CVE of the Week. We couldn’t do this without listeners like you.

Did we miss something? Do you have a topic you want us to cover?

Send feedback to feedback@itsparccast.com or reach out on social.

IT SPARC Cast

@ITSPARCCast on X

https://www.linkedin.com/company/sparc-sales/ on LinkedIn

John Barger

@john_Video on X

https://www.linkedin.com/in/johnbarger/ on LinkedIn

Lou Schmidt

@loudoggeek on X

https://www.linkedin.com/in/louis-schmidt-b102446/ on LinkedIn

Hosted on Acast. See acast.com/privacy for more information.

They also explore Google’s push to standardize AI-driven commerce through agent protocols and why Meta locking down more than 6 GW of nuclear power may be the clearest sign yet that energy—not silicon—is becoming the limiting factor for AI at scale. If you’re tracking AI strategy, platform lock-in, and the future of data centers, this episode connects the dots.

📌 Show Notes

00:00 – Intro

This week on IT SPARC Cast, John Barger and Lou Schmidt break down a week dominated by AI power shifts, enterprise pushback, and the growing reality that energy—not compute—may be the biggest constraint on AI’s future.

📰 News Bytes

00:52 – It’s Official: Apple Going with Google for AI

Apple confirms it will rely on Google’s Gemini models to power the next generation of Siri and Apple Intelligence. John and Lou discuss what this says about Apple’s AI strategy, the risks of deep vendor lock-in, and whether Apple can realistically switch models later without breaking workflows.

https://techcrunch.com/2026/01/12/googles-gemini-to-power-apples-ai-features-like-siri/

05:44 – Microsoft to Allow IT Admins to Uninstall Copilot

Microsoft is testing new Windows policies that allow enterprise IT teams to remove the consumer Copilot app from managed devices. The conversation explores enterprise data governance, Intune controls, and why this signals a broader shift toward AI choice rather than forced adoption.

https://www.bleepingcomputer.com/news/microsoft/microsoft-may-soon-allow-it-admins-to-uninstall-copilot-on-managed-devices/

09:46 – Google Announces a New Protocol for AI-Driven Commerce

Google introduces the Universal Commerce Protocol (UCP), an open standard designed to let AI agents handle shopping, payments, and transactions across retailers. With backing from major brands and payment networks, John and Lou unpack why agent-driven commerce may become one of AI’s first truly mainstream use cases.

https://techcrunch.com/2026/01/11/google-announces-a-new-protocol-to-facilitate-commerce-using-ai-agents/

12:47 – Meta Signs Nuclear Power Deals for AI Data Centers

Meta secures long-term nuclear power contracts totaling more than 6 GW to fuel its AI infrastructure. The discussion focuses on why power—not chips—is becoming the true bottleneck for AI expansion and why nuclear energy is rapidly moving from “controversial” to “necessary.”

https://techcrunch.com/2026/01/09/meta-signs-deals-with-three-nuclear-companies-for-6-plus-gw-of-power/

🔚 Wrap Up

16:49 – Mail Bag

Listener feedback revisits cross-platform AI agents, Apple’s closed ecosystem, and whether enterprises can afford to exclude Mac users as agentic AI becomes more central to daily workflows.

18:53 – Wrap Up

John and Lou close the episode by reinforcing a key theme: AI’s future will be defined as much by energy, policy, and interoperability as by model performance.

Hosted on Acast. See acast.com/privacy for more information.

The flaw is actively exploited in the wild, remains unpatched, and—ironically—uses the spam filtering engine itself as the attack vector. With no user interaction required and evidence of nation-state activity, this vulnerability represents a worst-case scenario for organizations relying on Cisco’s email security stack.

If you run Cisco Secure Email Gateway or Email Security Appliances, this is an emergency-level issue that demands immediate attention.

⸻

📌 Show Notes

🚨 CVE of the Week: CVE-2025-20393

•Severity: CVSS 10.0 (Critical)

•Status: Actively exploited, no patch available

•Vendor: Cisco

🎯 Affected Products

•Cisco Secure Email Gateway (SEG)

•Cisco Email Security Appliance (ESA)

•Cisco Secure Email and Web Manager (SEWM)

•All affected systems run Cisco AsyncOS

🔓 How the Exploit Works

•Attackers deliver a specially crafted email that is processed before a spam verdict is reached

•The payload is executed during email parsing, attachment handling, or content inspection

•No user interaction required

•The malicious email never needs to reach an inbox

💥 Real-World Impact

•Full remote code execution on the email gateway

•Email interception and exfiltration (espionage risk)

•Persistent access for follow-on attacks

•Credential harvesting and downstream phishing using trusted infrastructure

•Log wiping, making detection extremely difficult

🌍 Threat Activity

•Exploits observed as early as November 2025

•Linked to Chinese state-aligned actors

•Tracked under UAT-9686, associated with groups such as APT41 and UNC5174

•Added to CISA’s Known Exploited Vulnerabilities (KEV) catalog

🛡️ Mitigation Guidance (No Patch Available)

•Immediately restrict and segment management interfaces

•Tighten ACLs and allow lists

•Treat SEG as Tier-Zero-adjacent infrastructure

•If compromise is suspected: full system rebuild required

•Assume persistence due to log tampering

🧠 Commentary

•The exploit weaponizes the very system designed to stop malicious email

•Lack of a patch from a vendor of Cisco’s size raises serious concerns

•For some organizations, this may prompt reevaluation of email security platforms altogether

⸻

🔚 Wrap-Up & Listener Feedback

We want to thank listeners who continue to engage with the show and help shape the conversation:

•GFABasic32 wrote:

“Thanks for the emergency update on n8n. I love the balance of technical deep dives and high-level strategy. You guys make keeping up with CVEs actually entertaining.”

•Dennis added:

“I love the CVE of the Week. These episodes are like exposure therapy.”

That’s exactly the goal—helping you face what’s happening in security so you can respond, not react.

Have thoughts on this CVE or want us to cover another one? Reach out.

⸻

🔗 Social Links

IT SPARC Cast

@ITSPARCCast on X

https://www.linkedin.com/company/sparc-sales/ on LinkedIn

John Barger

@john_Video on X

https://www.linkedin.com/in/johnbarger/ on LinkedIn

Lou Schmidt

@loudoggeek on X

https://www.linkedin.com/in/louis-schmidt-b102446/ on LinkedIn

Hosted on Acast. See acast.com/privacy for more information.

From Intel’s attempt at a comeback, to AMD and Qualcomm’s positioning against NVIDIA, to growing concerns about trust, security, and AI agents living inside your OS, this episode separates meaningful signals from CES noise—and explains why power efficiency, autonomy, and control are becoming the real battlegrounds.

⸻

⏱️ Show Notes

00:00 – Intro

John and Lou frame CES as the unexpected epicenter of enterprise IT announcements, explaining why CPUs, AI, and robotics dominated the show—and why IT teams should care.

⸻

📰 News Bytes

00:54 – New CPUs Announced

CES saw major CPU launches from Intel, AMD, Qualcomm, and NVIDIA—signaling a shift toward mainstream AI hardware announcements. Intel launched Panther Lake, AMD expanded Ryzen AI, Qualcomm pushed Snapdragon X2 for AI agents, and NVIDIA moved Rubin into full production.

⸻

09:45 – Lenovo’s New AI Agent

Lenovo unveiled Qira, an agentic AI designed to work across PCs, phones, wearables, and enterprise systems alongside Microsoft Copilot. The move highlights a growing push toward cross-device AI coordination—and raises questions about Apple’s closed ecosystem.

⸻

12:40 – Microsoft Integrates Copilot Deep into Windows

Microsoft is embedding AI agent launchers directly into Windows, allowing third-party applications to register system-wide AI agents. While this may keep operating systems relevant, it introduces serious trust and security concerns around deep OS-level access.

https://blogs.windows.com/windows-insider/2025/12/19/announcing-windows-11-insider-preview-build-26220-7522-dev-beta-channels/

⸻

🔁 Wrap Up

19:03 – Mail Bag

Listener feedback sparks a discussion on cloud outages, cost structures, and whether on-prem alternatives are becoming viable again for certain businesses.

22:15 – Wrap Up

John and Lou emphasize that resilience in the cloud is still possible—but only if organizations are willing to pay for it—and invite listeners to share what CES announcements stood out to them.

IT SPARC Cast

@ITSPARCCast on X

https://www.linkedin.com/company/sparc-sales/ on LinkedIn

John Barger

@john_Video on X

https://www.linkedin.com/in/johnbarger/

Lou Schmidt

@loudoggeek on X

https://www.linkedin.com/in/louis-schmidt-b102446/

Hosted on Acast. See acast.com/privacy for more information.

They break down how a content-type confusion bug inside n8n’s webhook processing engine allows attackers to fully compromise systems, why automation platforms are uniquely dangerous when breached, and what this means for enterprises running self-hosted or lightly governed internal tooling. The episode also highlights listener feedback and calls out a community-built React security tool worth checking out.

⸻

Show Notes

CVE of the Week: n8n “Ni8mare” (CVE-2026-21858)

•What is n8n?

An open-source, self-hosted workflow automation platform similar to Zapier or Make, widely used in enterprise and regulated environments for visual API-driven automation.

•Severity & Scope

CVE-2026-21858 carries a CVSS 10.0, joining multiple recent n8n vulnerabilities rated 9.9–10.0. n8n has over 200,000 deployments across cloud and on-prem environments.

•Technical Root Cause

A content-type confusion flaw in webhook form-data handling allows attackers to bypass file validation and execute arbitrary code.

•Why This Is Dangerous

Workflow engines often touch identity systems, APIs, credentials, and business logic—making them high-value targets with blast radii far beyond a single server.

•Enterprise Takeaway

Shadow IT, internally built automation, and lightly governed enablement tools must be continuously audited. Patch known systems—and actively hunt for unknown ones.

https://thehackernews.com/2026/01/critical-n8n-vulnerability-cvss-100.html

Listener Highlight

Dennis called out the Ingram Micro ransomware outage, noting that he hadn’t realized just how disruptive that incident was. And he’s absolutely right—Ingram Micro going offline for roughly 9–10 days created a nightmare scenario for VARs, system integrators, and build shops that rely on Ingram for ordering, RMAs, and emergency drop-ship replacements.

To put the scale in perspective, Ingram Micro processes an estimated $30–40 million per day in transactions. Even if some revenue was recovered later, the operational disruption, reputational damage, and downstream impact across the supply chain were massive. This is exactly why incidents like this belong in the conversation when we talk about real-world IT security failures.

Thanks for the thoughtful comment, Dennis—we genuinely appreciate the feedback and the conversation it sparked.

Wrap Up & Community Engagement

This episode reinforces a core theme: automation without security oversight becomes an enterprise liability. IT teams must partner with business units—not just say “no”—while enforcing continuous audits and rapid patching.

Follow & Connect

IT SPARC Cast

@ITSPARCCast on X

https://www.linkedin.com/company/sparc-sales/

John Barger

@john_Video on X

https://www.linkedin.com/in/johnbarger/

Lou Schmidt

@loudoggeek on X

https://www.linkedin.com/in/louis-schmidt-b102446/

Hosted on Acast. See acast.com/privacy for more information.

From the deflation of the AI hype cycle and Apple’s inevitable AI acquisition, to quantum computing entering nation-state playbooks, nuclear power reshaping data centers, and lawsuits finally challenging cloud provider accountability, this episode puts both hosts on the record. At the end of the year, they’ll revisit every prediction and grade themselves—so these takes are meant to age in public.

⸻

⏱️ Show Notes

00:00 – Intro

John and Lou explain the format: ten total predictions for 2026, five each, shared live without coordination—and revisited at the end of the year for accountability.

⸻

🔮 2026 Predictions

01:09 – Lou: The AI Bubble Deflates

AI investment cools as rationalization sets in—money keeps flowing, but weaker players and inflated expectations begin to fall away instead of a full collapse.

01:29 – John: Apple Acquires an AI / LLM Company

Apple makes a major AI acquisition to avoid long-term dependence on competitors’ models and regain control over its AI strategy.

02:53 – Lou: AI Starts to Get Really Useful

AI shifts from hype to practical value, quietly improving everyday workflows and real-world systems rather than flashy demos.

04:11 – John: Nation States Use Quantum Computing

Evidence emerges that a nation-state is actively using quantum computing for espionage or cyber operations, even if never formally acknowledged.

04:45 – Lou: AI Sneaks Into Places We Never Expected

AI embeds itself into overlooked products and environments—especially AR, wearables, and location-aware systems—delivering small but meaningful gains.

05:50 – John: Negative Reaction to OpenAI Hardware

OpenAI’s hardware announcement is initially panned by the press and competitors, only to be vindicated later as its purpose becomes clear.

06:51 – Lou: Power Gets Real for Data Centers

Energy—not chips—becomes the primary constraint for cloud and enterprise infrastructure, forcing new generation strategies into production.

08:00 – John: Small Modular Nuclear Reactors Explode (In a Good Way)

SMRs rapidly gain funding, deployments, and valuations as they become the only scalable answer to data center power demand.

08:36 – Lou: The Privacy Environment Gets Weird

Geopolitics, AI agents, and shifting borders create inconsistent and unpredictable privacy regimes across regions.

10:11 – John: Lawsuits Over Cloud Outages

Major lawsuits—possibly class actions—emerge after cloud outages cause real-world harm, forcing legal accountability for uptime failures.

⸻

🔁 Wrap Up

11:58 – Wrap Up

John and Lou invite listeners to submit their own 2026 predictions and commit to revisiting all forecasts at year’s end to see who was right.

IT SPARC Cast

@ITSPARCCast on X

https://www.linkedin.com/company/sparc-sales/ on LinkedIn

John Barger

@john_Video on X

https://www.linkedin.com/in/johnbarger/

Lou Schmidt

@loudoggeek on X

https://www.linkedin.com/in/louis-schmidt-b102446/

Hosted on Acast. See acast.com/privacy for more information.

From long-lived remote code execution flaws in enterprise networking gear, to a ransomware attack that shut down a global distributor, to systemic cloud outages that shattered the concept of “five nines” availability, this episode looks at what really went wrong—and why it matters heading into 2026.

These weren’t theoretical risks. They were real-world failures that disrupted supply chains, exposed critical infrastructure, and forced the industry to rethink assumptions about resilience, cloud reliability, and operational security.

⸻

📋 Show Notes

🔥 Top 5 IT Security Fails of 2025

01:39 - #5 – Ruckus Networks

Ruckus suffered from multiple long-lived remote code execution and authentication bypass vulnerabilities that persisted across 2024 and 2025. Impacted products included SmartZone, ZoneDirector, Cloudpath, and ICX switch management interfaces. Several flaws allowed unauthenticated access to management planes, enabling attackers to take over wireless controllers, push malicious firmware, and pivot deeper into enterprise networks. The lack of timely patches and limited communication made remediation especially painful for customers.

04:32 - #4 – Ingram Micro

A ransomware attack forced one of the world’s largest technology distributors to effectively shut down operations for days. Ordering systems went offline, patch access was disrupted, and thousands of downstream partners and customers were impacted. While it remains unclear whether ransom was paid, the incident highlighted how a single distributor outage can cascade across the IT supply chain, delaying hardware replacements, breaking SLAs, and costing millions in lost revenue.

07:21 - #3 – SAP NetWeaver

CVE-2025-31324 exposed a critical unauthenticated remote code execution flaw in SAP NetWeaver’s Visual Composer. Actively exploited in the wild before many organizations were aware of its existence, the vulnerability gave attackers potential access to finance, HR, procurement, and supply-chain data. For enterprises running SAP at the core of operations, successful exploitation meant full application takeover and deep visibility into business processes.

10:26 - #2 – React

A severe remote code execution issue in React sent shockwaves through the software ecosystem. With an estimated one-third of cloud applications depending on React, attackers were able to chain exploits involving dependency poisoning, build pipeline compromise, and even client-side execution. While patches were released quickly, the sheer scale of affected deployments meant many systems remained vulnerable well after disclosure—and some still are.

12:23 - #1 – Cloud Outages

2025 marked the year that “five nines” effectively died. Major outages across AWS, Microsoft Azure, Google Cloud, Microsoft 365, and IBM Cloud caused multi-hour disruptions affecting identity systems, collaboration tools, healthcare platforms, and public-safety infrastructure. Many incidents were caused not by attackers, but by control plane failures, DNS issues, NTP misconfigurations, and cascading dependencies. The result: billions in estimated financial impact and renewed concern over life-critical workloads running entirely in the cloud.

Watch Cloud SLA Theater: Why 99.999% Uptime Is a Joke in 2025 - https://www.youtube.com/watch?v=ygcYoFBXdjQ

⸻

17:19 - Wrap Up

If you think we missed a major security failure—or disagree with our rankings—we want to hear from you. Reach out, leave a comment, or send us feedback. Your insights often shape future episodes.

🔗 Connect With Us

IT SPARC Cast

X: @ITSPARCCast

LinkedIn: https://www.linkedin.com/company/sparc-sales/

John Barger

X: @john_Video

LinkedIn: https://www.linkedin.com/in/johnbarger/

Lou Schmidt

X: @loudoggeek

LinkedIn: https://www.linkedin.com/in/louis-schmidt-b102446/

Hosted on Acast. See acast.com/privacy for more information.

⸻

📰 News Bytes

00:46 – CEOs Keep Spending on AI Despite Spotty Returns

Despite mixed financial outcomes, a growing number of CEOs plan to increase AI investment through 2026, viewing AI as strategically unavoidable rather than immediately profitable.

Key discussion points:

•Fewer than half of current AI projects are delivering clear ROI

•Strong gains in sales, marketing, customer service, and developer productivity

•Weak performance in regulated, high-risk areas like legal, HR, compliance, and cybersecurity

•Layoffs blamed on AI may result in long-term operational backlash

The hosts argue that AI should augment human expertise, not prematurely replace it—and warn against betting the company on incomplete automation strategies.

https://www.msn.com/en-us/technology/artificial-intelligence/ceos-to-keep-spending-on-ai-despite-spotty-returns/ar-AA1SkMcE

07:34 – 5D Glass Storage: Crystals for the Enterprise

A UK company, SPhotonix, is advancing 5D glass storage, capable of preserving data for billions of years by etching nanoscale structures into glass using femtosecond lasers.

Highlights include:

•360 TB per 5-inch glass disk

•Designed for permanent archival, not hot or warm storage

•Potential replacement for long-term tape archives

•Early write speeds are slow, but roadmap improvements are promising

This technology positions itself as a future-proof solution for enterprises, governments, universities, and cultural institutions facing long-term data retention challenges.

https://www.tomshardware.com/pc-components/storage/sphotonix-pushes-5d-glass-storage-toward-data-center-pilots

15:00 – Coursera Acquires Udemy for $930 Million

Online education giant Coursera is acquiring Udemy in a deal valued at approximately $930 million, creating a dominant force in enterprise and consumer e-learning.

Discussion points:

•Udemy’s strong practitioner-led course model

•Coursera’s academic and credentialing reach

•Expanded use of AI for assessments, personalization, and skills validation

•Potential shift toward a “market-driven university” model

The hosts see this consolidation as a net positive for enterprise IT teams responsible for compliance training, upskilling, and leadership development.

https://techcrunch.com/2025/12/17/coursera-and-udemy-enter-a-merger-agreement-valued-at-around-2-5b/

🔁 Wrap Up

20:00 – Listener Feedback

⭐ Community Call-Out: Abdullah’s React Audit Tool

A special shout-out to Abdullah ( https://x.com/ozkayabd ) who responded on X after a previous React CVE episode and shared an open-source tool to help teams audit their environments:

👉 React Audit Scanner

http://rsc-auditor.vercel.app

This tool allows teams to quickly check whether they may be impacted by recent React vulnerabilities. As always, review and validate any third-party tool before using it in production.

A special shout-out to Megan, who reached out after the episode with thoughtful feedback—and who’s doing important work to tackle a problem far too many people experience: ghosting of job applicants by recruiters and HR teams.

Megan is actively pushing for better communication, transparency, and basic professionalism in the hiring process. It’s a reminder that while we talk a lot about AI, automation, and efficiency, the human side of tech and hiring still matters. Follow her on LinkedIn:

https://www.linkedin.com/in/megan-juliano

Connect with the hosts and the show:

IT SPARC Cast

@ITSPARCCast on X

https://www.linkedin.com/company/sparc-sales/

John Barger

@john_Video on X

https://www.linkedin.com/in/johnbarger/

Lou Schmidt

@loudoggeek on X

https://www.linkedin.com/in/louis-schmidt-b102446/

Hosted on Acast. See acast.com/privacy for more information.

The discussion covers why RC4 persisted for so long, how legacy Active Directory and Kerberos environments kept it alive, and why attackers have continued to exploit it through techniques like Kerberoasting. The hosts also highlight the new logging, auditing, and PowerShell tools Microsoft released to help enterprises identify and eliminate lingering RC4 dependencies—without breaking production systems.

⸻

📋 Show Notes

🔐 Main Topic: Microsoft Disables RC4 by Default

•Microsoft is removing RC4 (Rivest Cipher 4) as a default cipher in Windows authentication after more than 25 years.

•RC4 has been known to be cryptographically broken for decades and has been actively exploited in real-world attacks.

•The change impacts Kerberos authentication across Windows Server 2008 and later.

•RC4 will still function only if explicitly re-enabled—which is strongly discouraged.

⚠️ Why RC4 Is Dangerous

•RC4 has been abused in Kerberoasting attacks against Active Directory environments.

•Weak encryption allows attackers to extract service account credentials offline.

•Keeping RC4 enabled significantly increases the blast radius of a compromised domain.

🛠️ What Microsoft Did Right This Time

•Added enhanced Kerberos logging (Event IDs 4768 and 4769) to identify RC4 usage.

•Released PowerShell scripts to audit domain controllers for RC4 dependencies.

•Published clear migration guidance to move environments to AES-SHA1 and stronger encryption.

•Provided visibility before enforcing the change, helping admins avoid outages.

🎧 Listener Feedback Highlight

•A YouTube listener praised the CVE of the Week format as being highly valuable from an ops and security standpoint.

•Strong validation that actionable vulnerability analysis resonates with enterprise IT teams.

⭐ Community Call-Out: Abdullah’s React Audit Tool

A special shout-out to Abdullah ( https://x.com/ozkayabd ) who responded on X after a previous React CVE episode and shared an open-source tool to help teams audit their environments:

👉 React Audit Scanner

http://rsc-auditor.vercel.app

This tool allows teams to quickly check whether they may be impacted by recent React vulnerabilities. As always, review and validate any third-party tool before using it in production.

⸻

🔚 Wrap Up & Social Links

IT SPARC Cast

@ITSPARCCast on X

https://www.linkedin.com/company/sparc-sales/ on LinkedIn

John Barger

@john_Video on X

https://www.linkedin.com/in/johnbarger/ on LinkedIn

Lou Schmidt

@loudoggeek on X

https://www.linkedin.com/in/louis-schmidt-b102446/ on LinkedIn

Hosted on Acast. See acast.com/privacy for more information.

From there, the show dives deep into the mounting backlash against U.S. data centers, with over 200 environmental groups demanding a halt to new builds—ironically accelerating plans for orbital data centers. The conversation then turns optimistic as the inventor of the Super Soaker unveils a breakthrough technology that converts waste heat directly into electricity, potentially reshaping geothermal and data center power economics.

Finally, the guys explore Boom Supersonic’s unexpected pivot—using jet engines as grid-scale power generators for data centers—and Google’s launch of managed MCP servers that allow AI agents to plug directly into core Google services with minimal integration effort.

⸻

⏱️ Show Notes

00:00 – Intro

This week: Google admits AI pain is coming, environmentalists push data centers toward orbit, waste heat becomes power, and AI agents get a universal plug.

⸻

HOT TAKE

00:55 – Google CEO on AI Layoffs: “We’re All Going to Have to Suffer Through It”

•Sundar Pichai acknowledges widespread layoffs and economic strain tied to AI adoption.

•John and Lou discuss why AI-driven efficiency gains are being used as justification for premature workforce cuts.

•Key argument: AI doesn’t replace people—it amplifies small teams and enables entrepreneurship.

https://www.msn.com/en-us/money/companies/google-ceo-says-we-re-all-going-to-have-to-suffer-through-it-as-ai-puts-society-through-the-woodchipper/ar-AA1S5Pzx 

⸻

NEWS BYTES

06:11 – More Than 200 Environmental Groups Demand Halt to New U.S. Data Centers

•Greenpeace and others cite water usage, power demand, and CO₂ emissions.

•~$64 billion in data center projects already delayed or halted.

•Lou explains why this pressure is accelerating interest in orbital data centers—one FCC license vs. hundreds of local permits.

https://www.theguardian.com/us-news/2025/dec/08/us-data-centers 

⸻

10:26 – Super Soaker Inventor Wants to Turn Waste Heat into Electricity

•Lonnie Johnson (inventor of the Super Soaker) unveils the Johnson Thermal Electrochemical Converter (JTEC).

•Works with small temperature differentials—no turbines, no moving parts.

•Could dramatically change how data centers source supplemental power.

https://www.ajc.com/business/2025/11/earth-needs-more-energy-atlantas-super-soaker-creator-may-have-a-solution/ 

⸻

13:08 – Boom Supersonic Uses Jet Engines to Power Data Centers

•Boom Supersonic repurposes its jet engine designs into natural gas turbines for data centers.

•Each turbine outputs ~42 MW; initial orders exceed 1.2 GW and are rapidly increasing.

•First deliveries expected in 2027; turbine factory opening next year.

•John and Lou connect this to job creation across manufacturing, operations, and IT management.

https://techcrunch.com/2025/12/10/google-is-going-all-in-on-mcp-servers-agent-ready-by-design/ 

⸻

16:44 – Google Launches Managed MCP Servers for AI Agents

•Google introduces managed Model Context Protocol (MCP) servers on GCP.

•MCP creates a universal “language” for AI agents to interact with tools and services.

•Reduces API complexity—ask questions, get results, take action.

•Free during public preview for enterprise customers.

•Lou calls this a major step toward AI-native enterprise workflows.

https://techcrunch.com/2025/12/10/google-is-going-all-in-on-mcp-servers-agent-ready-by-design/ 

⸻

Wrap Up

20:38 – Mail Bag & Wrap Up

•Listener feedback highlights interest in portable and containerized data centers.

IT SPARC Cast

@ITSPARCCast on X

https://www.linkedin.com/company/sparc-sales/ on LinkedIn

John Barger

@john_Video on X

https://www.linkedin.com/in/johnbarger/ on LinkedIn

Lou Schmidt

@loudoggeek on X

https://www.linkedin.com/in/louis-schmidt-b102446/ on LinkedIn

Hosted on Acast. See acast.com/privacy for more information.

With an estimated 33–35% of cloud-based services running React, attackers are already leveraging automated tooling to deploy cryptominers, Linux backdoors, and persistent malware across vulnerable systems. If you run React, Next.js, or containerized web workloads, this episode outlines exactly why this exploit is so dangerous, how attackers are weaponizing it, and what you must do right now to mitigate risk—from emergency patching to Zero Trust and micro-segmentation strategies.

⸻

Show Notes

🔴 CVE of the Week: CVE-2025-55182 (React Server Components RCE)

In this episode, John and Lou sound the alarm on a critical vulnerability in React Server Components that has escalated from disclosure to active, automated exploitation in the wild.

Key points covered:

•CVE-2025-55182 allows unauthenticated remote code execution via unsafe serialization and deserialization in React Server Component endpoints

•Vulnerable components include:

•react-server-dom-webpack

•react-server-dom-parcel

•react-server-dom-turbopack

•A related issue impacts Next.js App Router deployments, tracked separately as CVE-2025-66478

•Even applications that do not explicitly use server functions may still be exploitable if RSC support exists

🚨 Active Exploitation Confirmed

Lou shares real-time intelligence showing attackers using automated tooling dubbed “React-to-Shell”, delivering:

•Cryptocurrency miners

•Linux backdoors (PeerBlight)

•Reverse proxy tooling (CowTunnel)

•Go-based post-exploitation implants (ZinFoq)

This is no longer theoretical—production systems are being compromised right now.

🛡️ Immediate Mitigation Guidance

If you run React or Next.js workloads:

•Patch immediately to fixed versions

•Disable or strictly isolate RSC server function endpoints if not required

•Place RSC behind WAFs and strict network controls

•Harden container and OS permissions

•Implement payload anomaly detection

•Move toward micro-segmentation and Zero Trust architectures to limit blast radius

John and Lou emphasize that patching alone is no longer enough in an era of AI-accelerated exploitation.

⸻

Wrap Up & Community Feedback

The episode closes with listener feedback from LinkedIn discussing CXL memory pooling and how it is changing enterprise infrastructure economics—plus a recommendation to check out deep-dive demos from Serve The Home.

As always, the team invites listener input on whether future episodes should focus on individual CVEs or broader security themes.

⸻

Follow & Connect

IT SPARC Cast

@ITSPARCCast on X

https://www.linkedin.com/company/sparc-sales/

John Barger

@john_Video on X

https://www.linkedin.com/in/johnbarger/

Lou Schmidt

@loudoggeek on X

https://www.linkedin.com/in/louis-schmidt-b102446/

Hosted on Acast. See acast.com/privacy for more information.

This episode blends politics, enterprise IT strategy, security concerns, and cloud architecture trends—delivered with classic SPARC Cast sarcasm.

⏱️ Show Notes

00:00 – Intro

This week: Apple says goodbye to its AI chief, India tests mandatory surveillance apps, OpenAI hits the panic button, and Amazon+Google become “friends with benefits.”

NEWS BYTES

00:46 – Apple AI Chief Exits

Apple confirms that John Giannandrea, SVP of Machine Learning & AI Strategy, will step down in Spring 2026.

•He was Apple’s “big hire from Google” and led AI initiatives for eight years.

•His replacement: Amar Subramanya, reporting to Craig Federighi.

•John & Lou discuss Apple’s AI struggles:

– Apple Intelligence is “not what was promised”—delayed, underwhelming, and widely criticized.

https://www.apple.com/newsroom/2025/12/john-giannandrea-to-retire-from-apple/