What are the opening hours?

Monday 9am - 5pm Tuesday 9am - 5pm Wednesday 9am - 5pm Thursday 9am - 5pm Friday 9am - 5pm

Michael Plis l Cyberkite

Home
Australia
Melbourne, VIC
Michael Plis l Cyberkite

Follow me for Cybersec / AI / IT for small business. Showing usefulness & dangers of tech! IT Pro for over 18yrs. Contact via website www.cyberkite.com.au

Cyberkite - Book IT & Cybersecurity services in Aus for small business. Welcome to the official Facebook profile of Michael Plis, an experienced IT professional based in Australia. With over 18 years of experience in the field, Michael offers unique insights into the benefits and potential risks of technology from a neurodiverse perspective. As the founder of Cyberkite, Michael is committed to pro

viding smart IT and Cybersecurity services to small businesses across Australia. For more information on Cyberkite, please visit cyberkite.com.au/aboutus

You can also connect with Michael by following his LinkedIn profile for alerts, technology tips, and cybersecurity updates. You can also find him on LinkedIn, Twitter, Instagram by searching for "Michael Plis". If you're a small business in Australia, book a 30-minute IT & Cybersecurity consultation with Michael by visiting cyberkite.com.au/book. Additionally, if you're interested in having Michael speak at an IT or Cybersecurity event, please feel free to send him a direct message. Michael is a creative and passionate technology and cybersecurity professional. His exposure to computers and art at a young age, following his family's move from Poland to Melbourne, has fueled his passion for the field. He is also a strong supporter of mental health initiatives and believes in creating environments that promote good mental health.

What's unique about Cyberkite?

1. Our focus is exclusively on small businesses, so our services are tailored to meet their specific needs. We provide expert IT & Cybersecurity consulting and support services to help businesses stay safe and connected. We generally don't accept domestic customers.

2. We offer comprehensive IT & Cybersecurity purchasing advice that is well-matched to cover all of your organization's IT functions. This saves you time spent shopping around and ensures that you have everything you need to run your business efficiently and securely.

3. We can act as a trusted IT & Cybersecurity deal brokerage partner, saving you time and money on large orders with vendors.

4. We are vendor-neutral, which means we offer advice about IT products from several brands to help you get the right products for your small business.

5. We focus on providing quality services to minimize future IT issues. We also believe that choosing to buy the cheapest services and products now will cost you more in replacements down the road.

6. Our business is owned and operated by experienced IT professionals with over 40 years of experience helping small businesses select the right products and services.

7. We make the IT support experience easier by providing helpdesk email and portal for ongoing clients, as well as documentation to ensure you are always up-to-date. Book IT Support via cyberkite.com.au/support

8. We offer competitive service rates on a casual basis for Australian businesses for IT & Cybersecurity services. See service rates.

9. We are 100% Australian-owned, which means you are supporting an Australian business when you choose to work with us.

10. We are proactive and care about your experience with us. We communicate with you every step of the way through your IT journey.

11. We offer a range of financing and payment options to suit your changing circumstances, including buy now with 4 repayments, loans, direct debit for recurring payments, manual, and card options. Learn more: cyberkite.com.au/pay

Disclaimer: Please note that the opinions expressed by Michael on this platform are his own and may not necessarily reflect the views of Cyberkite. He shares his opinions based on his extensive experience in the IT and Cybersecurity industry, learning from the world's top subject matter experts and passing on this knowledge to his audience in the hopes of benefiting them. The information in all posts & articles is for information purposes only. This about section and all subsequent posts from March 2023 may or may not have been refined and edited with help of a generative ai service to help Michael articulate as he has a neurodiverse mind.

28/05/2026

The "Notice and Consent" illusion is officially dead. The OAIC’s 2026 Privacy Survey proves consumer trust requires deep operational change.

The Office of the Australian Information Commissioner (OAIC) just released its 2026 Australian Community Attitudes to Privacy Survey (ACAPS), and it is a massive wake-up call for anyone running a business.

The numbers are staggering, yet entirely predictable:

📊 The Concern Gap: 93% of people say protecting personal info is deeply important to them, and 87% are more concerned about privacy today than 5 years ago.

❌ The Illusion of Control: 78% feel they have little to no real control over how their data is handled. 68% say clicking "Consent" rarely or never feels like a genuine choice.

🛑 The AI Red Line: 93% state that using their personal data to train AI models is flat-out unfair and unreasonable. 71% reject letting an organization train an AI system on their data after a service relationship has ended.

⚖️ The Fairness Verdict: Only 10% of consumers believe that organisations' real-world data practices are actually fair.

THE GLOBAL CONTEXT:

This data directly mirrors what we are seeing across Europe and North America. Regulatory bodies globally (such as those enforcing the EU AI Act or U.S. state privacy mandates) are pivoting away from checkboxes. They are moving toward active "Privacy by Default" enforcement, testing whether your data infrastructure is fundamentally "fair and reasonable".

WHAT SMALL AND MID-SIZED BUSINESSES NEED TO DO TODAY:

Privacy isn't just an enterprise legal task anymore, it is your ultimate competitive advantage. 68% of consumers state they would actively choose digital services if they trusted their data was being handled responsibly.
Here is the common-sense roadmap to secure your client data by design:

🛡️ Enforce Privacy by Default: Ensure your software platforms default to the highest privacy settings automatically. Don't make clients dig through settings to opt out.

📉 Radical Data Minimisation: Stop collecting "just-in-case" data. If a piece of information isn’t strictly necessary to deliver your immediate service, do not ask for it.

🛑 Fence Your Data from AI Engines: If you plug customer data pipelines into generative AI extensions, ensure you opt out of data-sharing and model-training clauses. Your customers explicitly draw a red line here.

🗑️ Automated Lifecycle Deletion: Build structural lifecycle controls that securely purge client data once the service retention requirement has ended.

(I have attached the full infographic summary below so you can see the data breakdown at a glance! I have also provided links to the full report as well as a infographic PDF provided by them)

Do you think we need to increase privacy for client data? share your comments👇

27/05/2026

Microsoft just launched an autonomous tool to fight ransomware at machine speed. There is just one massive catch! 🪝

Yesterday, Microsoft previewed its new Automatic Device Isolation capability within Defender for Endpoint. The logic behind it makes total sense: modern malware moves so fast that human response times are effectively obsolete. By the time an IT analyst reads a red flag, files are already being encrypted.

Microsoft's solution is an AI-driven, automated "air gap" that cuts off a device the millisecond it detects an attack.

You bring an automated defense to an automated fight. Sounds perfect, right?

The Twist: The SANS Institute just published a warning paper on this exact feature. Their research indicates that if these autonomous settings aren't tuned perfectly, a clever attacker can leverage Microsoft's own automated defense tool to disable all user and administrator accounts across the network.

Essentially, the AI can be tricked into locking your security team out of the building while the hackers roam free in the background.

The Cyberkite Takeaway:

We are a big proponent of smart automation, but this highlights a critical rule of 2026 infrastructure engineering: Autonomous action tools are not "plug-and-play." If you are a mid-market business or under-resourced IT team relying on AI to act on your behalf, you cannot leave it unconfigured. Out-of-the-box automation without strict governance leaves a backdoor for hackers to turn your own weapons against you.

🔒 The Action Plan:

Tune, Don't Just Turn On: When deploying automated attack disruptions, carefully establish baseline exclusions for critical administrative and response accounts.

Never Rely on a Single Layer: Autonomous endpoint isolation is fantastic, but it must be backed by secondary out-of-band communication channels for your IT team.

The Human is Still the Pilot: AI can act as the first responder, but it should never have unchecked authority over core directory structures.

Are you ready to hand the keys to an autonomous AI system that can lock down your network instantly, or does this SANS warning make you want to keep your hands on the steering wheel?

Let's discuss. 👇

20/05/2026

Google I/O 2026 just wrapped up, and the direction is clear: We are officially entering the "Agentic AI Era."

If you thought Google was just building better chatbots, this year’s keynote proved their true goal is far bigger. Google is turning Gemini into an omnipresent intelligence layer running across all your apps, workflows, and devices.

The scale of this shift is staggering. Google revealed that it is now processing 3.2 quadrillion tokens per month across its AI products.

For IT providers and business leaders, here are the 5 massive takeaways from I/O 2026 that change how we design business tech:

⚡ 1. Gemini 3.5 Flash is Here: Google dropped their next-gen lightweight model. Operating 4x faster than rival frontier systems, it is designed specifically to handle complex, multi-step automated agent workflows without the latency bottlenecks of the past. It’s now the default powerhouse behind Google Search and Workspace.

👁️ 2. Gemini Omni (The "World Model"): Google DeepMind unveiled Omni. This isn't just a text-to-video tool. It natively understands real-world physics, gravity, and spatial interactions. It can ingest text, audio, images, and video simultaneously to generate context-aware, scientifically accurate multi-modal outputs.

🤖 3. Gemini Spark (Your 24/7 Personal Agent): This is where the cross-device vision comes alive. Spark is an autonomous background agent that acts on your behalf, handling your inbox, looking out for hidden fees, managing scheduling, and executing multi-app actions without waiting for a prompt.

🔍 4. The 25-Year Search Redesign: Google Search box is getting its biggest overhaul in a quarter-century. Powered by Google’s new Antigravity platform, Search will now continuously run background "information agents" to monitor topics for you and automatically build custom, persistent mini-dashboards right inside Chrome tabs.

👓 5. Seamless Smart Eyewear: Closing out the cross-device strategy, Google and Samsung officially showcased their upcoming "Intelligent Eyewear" glasses coming this fall, weaving spatial computing natively into the Gemini ecosystem.

The Bottom Line: Google’s goal isn’t an AI tool you open, it’s an AI system you live inside. For businesses, the infrastructure question is no longer "Which chatbot do we use?" but "How do we govern autonomous agents acting on behalf of our staff?"

What do you think of Google's new agent-first direction?
Are we ready to let AI run in the background?
Share comments 👇

14/05/2026

The end of the "Chatbot" era is here. Mira Murati’s Thinking Machines just redefined the AI interaction model.

If you’ve ever felt the awkward "turn-based" pause when talking to an AI, you know the bottleneck. Today’s models wait for you to finish, process your input, and then respond. It’s a rigid, sequential loop.

On 11 Mat 2026, Mira Murati (former OpenAI CTO) and her team at Thinking Machines Lab unveiled a research preview that fundamentally breaks this cycle.

What are Interaction Models?

Instead of "bolting on" real-time features through external software, Thinking Machines has built TML-Interaction-Small, a 276-billion parameter model trained from scratch to handle interaction natively.

WHY THIS CHANGES EVERYTHING FOR BUSINESS:

⚡ Micro-turns (200ms): The system processes conversation in tiny chunks, responding in under 0.4 seconds—significantly faster than Gemini Live or GPT-Realtime.

👂 Simultaneous Processing: The AI listens, sees your screen, and speaks at the same time. You can interrupt it, and it can proactively jump in to correct you as you work.

🧠 Dual-System Architecture: While the Interaction Model keeps the conversation fluid, a powerful "background" model handles deep reasoning and web searches asynchronously, weaving the results back in naturally.

🕒 Time Awareness: It has a built-in sense of elapsed time. It can notice if a process is taking longer than usual or proactively remind you of tasks without being prompted.

At Cyberkite, we focus on how AI boosts productivity for small businesses - this release from Thinking Machines could revolutionize desktop computing and mobile computing, and truly head towards AI computers that are voice operated and can see you. But also smarter humanoid robots will benefit from this.

This is the difference between a "tool" you have to command and a "collaborator" that stays in the flow with you.

Are you ready for AI that works with you, not just for you? 👇

13/05/2026

I just put Google Workspace’s "Gemini 3.1 Pro" and Anthropic’s Free "Claude" to the ULTIMATE SMB Test. The results were embarrassing! 😱

Google DeepMind has promised that their latest Gemini models can autonomously edit, format, and manipulate Google Sheets and Excel files. So, I gave it a real-world test: a messy timesheet file. It needed start/end times tightened to set rules, weekend entries removed, duration columns recalculated, and duplicates deleted.

Yesterday 12th May 2026, I ran this test on my paid Google Workspace subscription using Gemini 3.1 Pro.

The result? It completely froze my spreadsheet. It told me it was "unable to do a complex set of editing actions." I tried it in the app. I tried the side panel. It sampled 5 rows and then gave up. I am paying for an ecosystem that promises autonomous productivity, but forces me to jump through a hundred fragmented apps just to get a failure message.

Then, I took the exact same messy Excel file and instructions over to the Free version of Anthropic Claude.

What happened next BLEW my mind.

Within minutes, Claude analyzed the instructions and realized a chatbot couldn't just "guess" the data. So, it autonomously wrote mini Python scripts in the background to execute the logic. In 10 minutes, the file was perfectly cleaned. It even generated a built-in preview window highlighting the removed weekend entries in yellow. I spent another 20 minutes prompting it to refine the rules, and it updated perfectly every single time. 30 minutes. Zero freezes. A perfect file. ON A FREE ACCOUNT. Time saved? At least half a day of work normally or more because I'm really slow with this stuff.

THE REALITY CHECK FOR BUSINESS AI:

Google is failing because they are trying to force AI to act like a traditional spreadsheet menu. Anthropic is winning because they have given Claude a secure sandbox to actually write code and do the work for you, all in one window. Question I have for Demis Hassabis: is there any plan to get ahead of Claude?

With the new "Claude Skills Marketplace" allowing businesses to add custom tools and workflows to their accounts, Anthropic isn't just beating Google's spreadsheet tools, they are replacing the need for them entirely.

Are you sticking with Google Workspace for AI, or are you migrating your heavy-lifting to Claude? 👇

08/05/2026

🚨 IMPORTANT UPDATE FOR STUDENTS & PARENTS 🚨

If your school, college, or university uses Canvas to submit assignments and check grades, you need to read this! Starting on May 7, 2026, the platform suffered a massive global cyberattack and remains offline today, May 8.

Here are the facts:

🧑‍🎓 Who is affected? Up to 275 million students and teachers across 9,000 schools globally.

🔓 What was stolen? A hacking group called ShinyHunters claims to have stolen student/staff names, email addresses, school ID numbers, and messages sent on the platform. (Thankfully, authorities say no financial information or official ID documents seem to be compromised yet).

🏫 How is education impacted? Online learning services are heavily disrupted right now, but officials note that physical schools and universities remain open.

⚠️ WARNING: The hackers are threatening to leak the data by May 12, 2026, unless schools pay a ransom. Because emails and names were leaked, please be on high alert for scams. Do NOT click on random links in texts or emails claiming to be from your school or Canvas, as criminals use this data to trick people!

Stay safe and check with your specific school for updates on how classes will proceed heading into the weekend! 📚🛡️

06/05/2026

The AI Arms Race just became a literal Arms Race. 🛑🤖

In the last 48 hours, the entire landscape of AI and Cybersecurity shifted. Here is what you need to know:

1️⃣ The Pentagon Deal: The US Military just signed agreements with OpenAI, Google, Microsoft, and Nvidia to integrate AI into classified military operations.

2️⃣ The Hacking Panic: Last month, a new AI model called "Claude Mythos" proved it could autonomously hack networks. It scared the US Government so badly that yesterday, they announced they will start pre-testing all future AI models before the public gets them.

3️⃣ The Business Reality: The World Economic Forum just reported that AI is now the "defining force" in cybersecurity.

If the government is panicking about AI hackers, small and medium businesses need to wake up. The tools hackers use are getting cheaper, faster, and autonomous.

You don't need to be a military contractor to be a target. Get your IT fundamentals right: Patch your software, use active threat detection, and lock down your network.

👇 Are we moving too fast with AI? Let us know your thoughts!

28/04/2026

Google DeepMind just killed the dream of Conscious AI. 🛑

If you’ve been worried that Artificial Intelligence is going to "wake up," become sentient, and turn into the Terminator, you can officially take a breath.

A Senior Scientist at Google DeepMind just published a massive paper proving that AI consciousness is structurally impossible. He calls the belief that AI can wake up the "Abstraction Fallacy." Here is the easiest way to understand it:

Imagine using a powerful computer to perfectly simulate a plant undergoing photosynthesis. The computer can track every atom and every light beam. But no matter how perfect the simulation is, the computer will never actually produce real glucose. Simulation is not reality.

DeepMind argues that code is just symbols. For code to mean anything, a conscious human has to observe it and assign it meaning. Without humans, AI is just electrical voltage moving across silicon. It doesn't know anything.

This is huge for the business world. It means we can stop worrying about science fiction, and get back to treating AI like a highly advanced hammer, a tool that needs strict governance, privacy controls, and human oversight.

Do you buy DeepMind's argument? Or do you think an AI could eventually "feel" something?

27/04/2026

Is Apple about to become just a "white-label" hardware brand?

With the massive news that Tim Cook is stepping down and handing the reins to hardware leader John Ternus, everyone is debating Apple’s AI strategy. A lot of people are praising Apple for refusing to build their own AI "engine," opting instead to act as the "dashboard" and rent AI from Google or OpenAI.

I strongly disagree. I think Apple has had its head in the sand for a long time, like an emu.

Think about the car industry: we are seeing a shift where tech giants build the chassis (the phone) but outsource the engine (the AI). While that saves money upfront, it creates massive Integration Debt. Apple's historic success comes from owning everything, the hardware, the software, and the silicon.

More importantly, some people are comparing AI to electricity, saying Apple doesn't need to own the power plant. But AI isn't electricity; it is the intelligence and data-processing layer of your entire digital life. Apple's biggest selling point has always been absolute user privacy and a closed ecosystem. If you outsource the "brain" of the iPhone to an external provider, you instantly compromise that closed-loop security.

They don't need to buy a power company, but they absolutely need to own the brain of their devices. By outsourcing it, they are selling their core philosophy short.

Do you think Apple can maintain its strict privacy standards while relying on third-party AI, or is this the beginning of the end for their closed ecosystem?

23/04/2026

Google just declared the end of the standalone AI chatbot and the dawn of the "Agentic Enterprise" era.

For the past couple of years, businesses have been dipping their toes into AI—using it to write emails, draft marketing copy, or brainstorm ideas. But behind the scenes, the tech giants are moving lightyears ahead.

At their latest Cloud Next convention, Google revealed that 75% of all new code written at their company is now generated by AI. Their engineers aren't just coding anymore; they are directing fully autonomous "digital task forces" to complete complex migrations in a fraction of the time.

This is no longer about human employees using a smart tool. It is about businesses deploying thousands of autonomous AI "agents" that work around the clock.

To support this massive shift, Google is launching a new Agent Platform to act as "mission control" for these digital workers, rolling out heavily upgraded cybersecurity defenses in partnership with Wiz, and deploying next-generation computer chips specifically designed to run millions of agents at once.

The question for business leaders and IT teams isn't whether AI will change your workflow—it's whether your current systems are secure enough to handle an entirely autonomous digital workforce.

Is your business actively planning for how autonomous AI agents will change your daily operations?

(Link to the official Google announcement is in the comments!)

Address

Melbourne, VIC

Opening Hours

Monday	9am - 5pm
Tuesday	9am - 5pm
Wednesday	9am - 5pm
Thursday	9am - 5pm
Friday	9am - 5pm

Website

https://cyberkite.com.au/, https://blog.cyberkite.com.au/, https://cyberkite.com.au/book

Alerts

Be the first to know and let us send you an email when Michael Plis l Cyberkite posts news and promotions. Your email address will not be used for any other purpose, and you can unsubscribe at any time.