ITCloudLink LLC - Managed Computing Services

10/09/2026

Someone at your front desk has a client letter open in one window and a chatbot open in the other. They ask whether it's okay to paste one into the other.

Who in your office answers that?

If nobody does, the fix is five decisions, and none of them is a purchase:

1. Name one person who can say yes, no, and stop.
2. List every AI tool already in use. Give staff two weeks.
3. Sort the list by what a wrong answer would cost.
4. Write one page: which tools, what information, what's off limits.
5. Put the next review on the calendar.

All five, in plain English:
https://itcloudlink.com/blog/five-decisions-before-your-office-buys-any-ai/

08/09/2026

Three posts, one argument:

1. Fraud enters through your business, not the bank.
2. Insurance pays after the loss. It doesn't close the entry point.
3. The tools to close it already exist. What's missing is a perimeter that reaches your door.

This last post gets specific: what endpoint defense actually looks like for a business that moves money, why the cloud shared-responsibility model is the right analogy, and what we do at that layer.

Some of it is infrastructure you'll never touch. But the practical version is available right now, and the first step is smaller than most people expect: stop routing MFA through SMS on any account with financial access.

Law firms. Medical practices. Mid-market operators. If you're in LA and you handle client funds or regulated records, this one's for you.

Link in the comments.

06/29/2026

☔️When I ask clients how they've handled their cybersecurity exposure, the most common answer is "we have insurance."

That's a real answer. It's not the complete one.

The UK tried the logical version of this: mandate that banks reimburse victims of authorized fraud. The banks paid. The fraud rose anyway. Moving who pays doesn't change where fraud enters the system.

Cyber insurance works the same way. It pays out after a loss. It doesn't close the entry point.

New post on the IT CloudLink blog goes through the UK data and what it means for how SMBs should actually think about fraud protection -- especially law firms and medical practices where a breach carries regulatory exposure that no policy covers.

Link in the comments. -

06/27/2026

The AI tools that could help the most with your real work are the ones you can’t use for it.

UpToDate Expert AI. Westlaw CoCounsel. ChatGPT. All cloud products. All require regulated data to leave your building the moment a patient or client enters the conversation.

OCR’s December 2024 Security Rule update made it sharper: AI systems now have to appear in your mandatory technology asset inventory, and staff using consumer tools without IT oversight is an explicit enforcement category.

There’s a configuration that resolves this – not through a better contract with a cloud vendor, but because the hardware is physically on-site and the data never crosses a network boundary.

New post on the IT CloudLink blog explains what that looks like for a 10-20 seat medical practice or law firm, and what it costs.

Link in the comments. -

06/22/2026

One of the hardest conversations I've had with a client is explaining that the nearly half million dollars they just lost isn't coming back -- and the bank isn't liable.

The fraud came through their email. It looked like an invoice from a vendor they'd worked with for years. By the time anyone noticed, the transfer was done.

The bank's fraud systems didn't fail. Those systems are designed to protect the bank's perimeter. Your email inbox is not inside that perimeter.

New post on the IT CloudLink blog breaks down exactly where business fraud enters the financial system -- and why knowing this matters more than your current security setup might reflect.

Link in the comments. -

05/12/2026

A byline I wrote has been published today in Fintech Bloom: "The Fraud Liability Debate Is Asking the Wrong Question."

The piece draws on direct incident experience -- two cases where fraud entered through an SMB client endpoint and flowed through financial infrastructure. Different outcomes on liability. Same undefended entry point.

The argument: financial services has extended its revenue surface to the SMB endpoint. Its security posture hasn't followed. Shifting liability doesn't solve that.

If you work in fintech, banking, compliance, or SMB IT -- this one's worth a read.

Link in comments.

05/08/2026

https://www.youtube.com/live/M8VYWQTCzCA?si=vfPiG-dZfEIYVgQB

David Edward of Abacus joins us to discuss the rise of private AI infrastructure, on-prem AI servers, and why regulated industries are moving away from the c...

05/08/2026

SpaceXAI will provide Anthropic with access to Colossus 1, one of the world’s largest and fastest-deployed AI supercomputers, to provide additional capacity for Claude → http://x.ai/news/anthropic-compute-partnership

𝕏 xAI

05/03/2026

🔥 The train keeps moving... It's built on a simple idea: I've spent 30+ years building and securing IT infrastructure for healthcare, legal, and enterprise clients. Now I'm applying that same thinking to private AI -- and putting it on camera. 🎭One episode already in the can. Next one scheduled.🔗 [rondthaler.dev](http://rondthaler.dev) 🌤

04/13/2026

🔥Something new just went live - our AI division! 👋

I've spent years inside some serious IT environments — UCLA Health, TV Guide, Fox Sports, Toyota Information Systems/HQ — and now I'm putting all of that to work for small and mid-sized businesses that need real AI infrastructure, not cloud subscriptions.

rondthaler.dev is live. Private AI systems. Local, secure, and built for businesses that handle sensitive data.

Come take a look. 👉 rondthaler.dev