Derek Gabriel

Founder & CEO of @igniteSG; Aloha! I'm on a mission to eradicate small business cyber security risks. I write and speak on cybersecurity and technology.

[Derek Gabriel @dsghi]

AI adoption isn’t splitting businesses into “tech-savvy” and “anti-tech”.

It’s splitting them into those moving at different speeds inside the same company 😬

New research shows a big age-related gap in how people use AI at work.

Roughly half of under-35s are already using AI tools regularly. Many have had training. Most see AI as helpful for their jobs.

But around half of over-45s haven’t used AI at all.

Not because they don’t trust it or because they think it’s dangerous.

Mainly because it feels unfamiliar.

And that’s where the real risk sits ⚠️

When adoption is uneven, AI goes underground.

Some staff quietly use AI to move faster. Others avoid it completely.

Managers assume “we’re not really using AI yet”, when really, parts of the business already are.

That creates problems like inconsistent outputs, unclear data handling and no shared standards.

And of course, no confidence about what information is being fed into which tools.

The research also highlights something important: Countries and organizations with slower, more cautious adoption aren’t falling behind because of a lack of tools. They’re falling behind because of a lack of confidence and guidance 🤷‍♂️

AI doesn’t need to be everywhere to be useful. But it does need to be understood.

The businesses that get the most value won’t be the ones chasing every new AI feature.

They’ll be the ones that:

�� Set clear boundaries
• Give people simple, practical training
• And focus on using AI to remove friction, not create anxiety

❓ Is AI in your business something you’ve consciously decided how to use, or is it being used quietly, inconsistently, and without a plan?

[Derek Gabriel @dsghi]

There’s a lot of noise about AI right now, but this caught my eye because it’s refreshingly honest 🙂

A report shows that around 70% of retailers are already testing or partially using agentic AI.

But only 8% have rolled it out fully across their business.

In other words, most people are experimenting. Very few have cracked it.

And I’m certain it doesn’t only apply to retail.

Agentic AI isn’t just a chatbot answering questions.

It’s AI that can look across systems, spot issues, and suggest (or trigger) actions. Think delays, bottlenecks, stock problems, or inefficiencies. Not marketing slogans.

Retailers are optimistic.

Nearly all believe AI will be essential to staying competitive, and many expect efficiency gains very soon.

But they’re also hitting reality.

The biggest blockers?

• Data that isn’t clean or joined up
• Concerns about trust, transparency, and regulation
• And a shortage of people who know how to implement AI properly

What’s interesting is where AI is heading.

So far, most use has been in customer service and marketing.

But the next wave is about operations. Things like inventory, supply chains, fulfilment, admin. Less creative AI, more quietly fixing problems before customers notice.

And that’s the bit business owners should pay attention to.

The real value of AI is removing friction from day-to-day operations and freeing humans to focus on decisions that need judgment.

AI works best when the foundations are solid: Good data, clear processes, and realistic expectations.

So, here’s my question for you 🤔 If AI could spot problems in your operations before they became issues, would your systems be ready to support it?

[Derek Gabriel @dsghi]

Dropped into a long email chain halfway through? Catch up without reading every reply from the top…

[Derek Gabriel @dsghi]

This is why phishing is getting harder to spot… and why “just be careful with emails” isn’t enough anymore 😬

Attackers are abusing legitimate Google services to send phishing emails that look completely genuine.

Not fake domains or sender addresses.

Actual emails coming from Google-owned infrastructure.

Security researchers recently tracked almost 10,000 phishing emails sent to thousands of businesses in just two weeks.

The emails looked like standard Google notifications. Voicemails, shared documents, that sort of thing.

And they were sent from a real @google .com address.

This wasn’t Google being hacked.

Attackers were misusing a real Google Cloud automation tool to send emails as part of a workflow.

Because those emails are generated by Google systems, they inherit Google’s excellent sender reputation.

That’s what makes them so dangerous.

Clicking the link takes victims on a journey that feels safe at every step:

• A trusted Google Cloud link
• A convincing “prove you’re human” CAPTCHA
• Then… a fake Microsoft login page

By the time someone realizes what’s happened, their email password has already been handed over.

Most of the victims were in manufacturing, tech, and finance. But the technique itself isn’t industry specific.

If your business uses Microsoft 365 and trusts Google links, this applies to you too.

The takeaway is simple but important: You can’t rely on trusted brands as a safety check anymore.

And you can’t expect staff to spot every trick.

That’s why modern security focuses on layers. Things like multi-factor authentication, conditional access, and reducing what a stolen password can do.

Because today’s phishing looks normal.

👉 If one of your team received a genuine-looking email from Google, would they act on the request without thinking?

[Derek Gabriel @dsghi]

This is one of the more worrying phone scams I’ve seen recently, because it looks completely legitimate 😬

Researchers have uncovered a campaign where attackers are taking real banking and government apps, modifying them with hidden malware, and then tricking people into installing them.

These are not fake apps created from scratch.

They’re actual apps, poisoned ☠️

A person receives a text or email that looks like it’s from a trusted organization. Maybe a power company. Maybe a government department. Sometimes it escalates into a phone call “to help”.

They’re told they need to make a payment or resolve an issue urgently.

They’re then directed to a very convincing website, sometimes it’s even one that looks like the official app store, and asked to download an app.

The app behaves exactly like the real one.

But behind the scenes, it’s been altered.

Once installed, the app asks for permissions it doesn’t really need. If the user agrees, attackers can:

🚫 Steal login details
🚫 Commit banking fraud
🚫 Monitor activity
🚫 And in some cases, take full control of the device

The scariest part?

The malware can clean up after itself.

Victims may never realize what happened until money disappears or accounts are accessed.

So far, this campaign has mainly targeted parts of Southeast Asia, but the technique itself could work anywhere.

And the rule that protects you hasn’t changed: Legitimate banks and government departments DO NOT ask you to install apps via text messages, links, or phone calls.

If something arrives unexpectedly and creates urgency, pause. Don’t click. Don’t download. Verify independently.

Because once a malicious app is installed, the damage is already done.

👉 If a message claimed to be from a trusted authority and asked you to install an app urgently would you naturally stop and verify or feel pressured to act?

[Derek Gabriel @dsghi]

If you’re honest, you may say some of the AI stuff being pushed into Windows 11 lately has been annoying 😅

If there are features you won’t use, it can feel like Microsoft is adding noise rather than value.

But buried under all that hype, there are a few AI ideas that really do make sense.

One of the best features Windows ever added was Clipboard History.

Press Windows key + V and you can see everything you’ve copied recently, not just the last thing. Once you get used to it, it’s hard to live without.

Now imagine instead of pasting what you copied, Windows could help you reshape it.

Copied a block of text? Turn it into bullet points.

Copied messy data? Convert it into a table.

Copied an image? Remove the background before pasting.

All right from the clipboard, without opening another app.

That’s what Microsoft’s experimenting with.

It’s patented an idea that essentially adds “advanced paste” to the Windows clipboard, using AI to offer smart options when you paste.

That saves you the boring formatting steps.

This is the kind of AI I like.

Small, practical help where it already fits naturally into how we work.

It’s early, and it’s not built into Windows yet, but it shows where Microsoft could be heading.

And honestly, if AI is going to win people over, it’ll be through features like this. Quiet. Optional. Genuinely useful.

❓Would you be more open to AI in Windows if it only showed up when it made your work easier?

[Derek Gabriel @dsghi]

This one catches people out because it uses something we all trust without thinking about it 😬

Your calendar 🗓️

Security researchers are warning that a perfectly normal calendar feature can be abused to deliver phishing links and scams, without ever sending an email.

Most calendar apps let you subscribe to external calendars. Think public holidays, school terms, sporting events, industry events, company schedules… once you subscribe, events just appear automatically in your diary.

That’s convenient.

And that’s the problem.

If the organization behind that calendar shuts down, or lets their website domain expire, the calendar subscription doesn’t magically stop.

It keeps trusting whatever lives at that address.

If a cybercriminal later buys that expired domain, they can start pushing new events straight into people’s calendars.

Those events can contain links to fake login pages, urgent notices, or malware downloads. And because they appear in your calendar, they feel legitimate.

Researchers found hundreds of abandoned calendar domains still feeding events into people’s devices.

Some were for public holidays, religious calendars, even major sporting events.

Altogether, they estimate around four million devices are affected. And that’s likely an understatement.

This isn’t a bug. Your calendar app isn’t broken. It’s doing exactly what it was designed to do: Trust subscribed calendars.

The risk comes from subscriptions being forgotten about and never reviewed.

From a business point of view, this matters more than you might think.

People trust their calendars. If something appears there, it feels official. And attackers know that.

The fix isn’t technical. It’s behavioral.

Every so often, it’s worth asking:

• What calendars am I subscribed to?
• Do I still recognize and trust the source?
• Do I need this anymore?

Removing old subscriptions reduces risk without affecting how you work day to day.

👉 If something suspicious popped up in your calendar tomorrow, would you question it or assume it must be safe because it’s in the diary?

[Derek Gabriel @dsghi]

Every quarter the same analyst note circulates: NVIDIA's revenue is circular. The money just goes in circles.
I'm a Microsoft channel partner. I just watched GTC. They're wrong - and Jensen told everyone exactly why on Monday.
https://t.co/nKryMcn8QT

[Derek Gabriel @dsghi]

Automate ONE workflow - learn how at the @KailuaChamber AI Hui Lunch & Learn with Liza Magcale & Heather Di Rocco.
🗓️ Nov 10 11:30–1 Kailua Lunch included!
🔗 https://t.co/sfbQF9GnVK #AI #Automation #HawaiiBusiness

[Derek Gabriel @dsghi]

@DivesTech You're an absolute moron for believing him at this point. I guess it's why you wear the clown suits. Any other company reporting earnings and financials like this would have it's value cut in half. Elon hasn't told a true about Tesla yet.

[Derek Gabriel @dsghi]

Total buying opportunity. Whenever someone compares this to the dotcom bubble, I roll my eyes. No business is ditching their AI systems for Chinese models. Even if the claims are true, competition accelerates research and innovation spending.

[Derek Gabriel @dsghi]

Join the waitlist for Flic Duo and be the first to experience the future of smart home control. #FlicDuo #SmartHome https://t.co/IaZHPbHQBv (What a fun marketing campaign)

[Drone Advocacy Alliance @DAA_Drones]

Too often, the voices of drone users are drowned out in policy discussions that threaten to restrict the ability of first responders, researchers, and recreational flyers to choose and use the best drones for their work. Take action: https://t.co/K2vAex1lbB

[Derek Gabriel @dsghi]

@teslaenergy Oh, they left off that their solar panels are made in China...

[Derek Gabriel @dsghi]

Why do Uber drivers insist on calling, the whole f-ing reason Uber exists is the app 🤦🏼‍♂️

[Paul Couvert @itsPaulAi]

Microsoft has just announced Copilot's ability to process data in Excel by generating Python code.

- Request a detailed analysis
- Copilot generates Python
- Excel executes the code to show the result

All without having to write a single formula. Just natural language.

Huge deal for data analysis and visualization.

[CIRA @CIRA_CSU]

Three storms in the central and east Pacific Ocean participate in a conga line.

Newly-formed Tropical Storm Hector and Hurricane Gilma churn over open waters while Tropical Storm Hone moves to the west of Hawaii.

[Derek Gabriel @dsghi]

@hulu_support @SOGamingYT The problem is you sell me 'NO ADS,' and I still get ads; it seems like fine print BS, and you should be fined for false advertising. If it's limited ads, call it that. YOU named it 'no ads'.....

[Derek Gabriel @dsghi]

I just backed American Peaks Crystal Whiskey Set on @Kickstarter https://t.co/zk0gyo439e

[Derek Gabriel @dsghi]

It's funny how people would rather take a pill than change their diet. We toyed with this over a decade ago in Honolulu. There was not as much support for the idea as I had expected, though things have changed significantly since then, including the global pandemic, GLP-1s, AI, etc.